risk tolerance essay

How Risk-Tolerant Are You?

In how many of these 13 scenarios would you take the risk.

Posted September 9, 2021 | Reviewed by Abigail Fagan

• Considering your risk tolerance in different scenarios can help you know yourself better and improve your decision-making.
• Risk tolerance can be consistent or vary across work, relationship, health, financial, and recreational domains.
• Considering the opportunity costs of each dilemma can help you decide how to move forward.

In Ruth Rendell’s novel, Kissing the Gunner’s Daughter, an off-duty policeman had just taken his teenage son’s authentic-looking toy revolver from him, and the cop put it in his own pocket. The policeman went to the bank to make a deposit whereupon a man and a boy rob the bank. The man shoots and kills the security guard and runs out, leaving the boy with what the cop assumes is a toy gun. The cop decides to take the risk of pulling out his son's toy revolver. He points it at the kid and orders, “Drop the gun!” The kid’s gun is no toy and he shoots the cop dead.

Turning to the real world, we're all faced with risky decisions. For each of these 13 situations, ask yourself if you'd take the risk. Your answers will help you clarify your risk tolerance: how much risk you’re willing to take regarding your work and personal life.

As you're deciding, consider what, in your case, would be the opportunity cost of making the risky choice. For example, someone who is relatively satisfied in their career would have more to lose in pursuing their dream career than would someone who hates his or her career.

You’ve always wanted a career in the performing or visual arts and been praised by teachers, friends, and family: for example, “You really should try to be a professional singer!” But you’re now 25 and gotten only $50 or $100 gigs: total annual income $2,000. Total annual expenses for lessons, costumes, and travel: $3,000. Do you press on?

You see good burrito shops quite crowded. Contemplating opening your own, you work in one to learn the trade. Now, you think you have a good sense of what to do although you worry about raising enough money, hiring well, quality control, insurance, and complying with state and local government regulations. The other alternative you’re considering is a low-risk one: a job in accounting. Which do you choose?

You see a cool job opening but meet only two-thirds of the requirements and sense that your chance of landing the job, even if you create a great application and prepare well for the interviews, is only 5 or 10 percent. Do you apply?

You’d love to work for Apple but there’s no appropriate job opening. So, on LinkedIn, you find the names of a few Apple employees whose job title suggests they’d have the power to hire you. You know that cold contact only occasionally results in a job offer. Do you take the risk of contacting them?

Your boss doesn’t like an idea that you think is great. Do you take the risk of bringing it up to someone else, perhaps your boss’s boss or at a staff meeting, or do you let it go?

You hate a component of your company’s training program but you saw someone else complain about it and soon was demoted. Do you complain?

• Relationships

You have your romantic eye on someone who is more attractive, has a better personality , and a better job than yours. Do you ask the person out for coffee?

You’ve been married for a few years and seen the relationship decline from B+ to C. You like being in a couple and think there’s a significant chance you won’t find anyone better. Do you leave the marriage ?

You’ve grown disgusted with your adult child: meanness, laziness, lousy friends, drug abuse . And that’s despite your diligent efforts to be a good parent. Now, your child has just dropped out of college and after a desultory job search says, “I can’t find a job. I want to come and live back home.” The chances of that ultimately benefiting your child, let alone you, are small. Do you say yes?

You’re 20 pounds overweight and your doctor said, “You should lose it.” But you know that the vast majority of dieters, despite the months-long sacrifice, gain back the weight. Do you commit to losing the weight and keeping it off?

You’ve become dependent on your psychotherapist but you’re revisiting the same issues again and again and your life isn’t much better. Alas, you’re pretty sure that if you discontinue, you’ll be worse and, you don’t want to offend the therapist who has tried so hard to help you. Do you quit?

Your friend is a mortgage broker who tells you that you can get a 10 percent annual return by lending on a piece of real estate that has plenty of equity in it. But you heard about a similar situation in which the appraisal was phony and so, in fact, there was no equity in the property and the owner ran away without paying the mortgage holders. You ask your friend, “If the loan is so secure, why doesn’t a bank make the loan?" Your friend replies that the owner said that the banks wanted to charge 12 percent. Do you make the loan?

You have long planned a trip to Europe and finally saved up enough to afford low-season prices: mid-November. Even though you’re vaccinated, you’re nervous about going, both because of COVID and because you’d hate if you got the snowy, cold weather that often but not always occurs at that time of year. Do you go?

The takeaway

People vary on all sorts of factors. Risk tolerance is perhaps an under-considered one. Have your answers to these questions clarified anything about your risk tolerance? As a result, is there anything you want to do differently?

I read this aloud on YouTube.

Marty Nemko, Ph.D ., is a career and personal coach based in Oakland, California, and the author of 10 books.

• Find a Therapist
• Find a Treatment Center
• Find a Psychiatrist
• Find a Support Group
• Find Teletherapy
• United States
• Brooklyn, NY
• Chicago, IL
• Houston, TX
• Los Angeles, CA
• New York, NY
• Portland, OR
• San Diego, CA
• San Francisco, CA
• Seattle, WA
• Washington, DC
• Asperger's
• Bipolar Disorder
• Chronic Pain
• Eating Disorders
• Passive Aggression
• Personality
• Goal Setting
• Positive Psychology
• Stopping Smoking
• Low Sexual Desire
• Child Development
• Therapy Center NEW
• Diagnosis Dictionary
• Types of Therapy

Understanding what emotional intelligence looks like and the steps needed to improve it could light a path to a more emotionally adept world.

• Coronavirus Disease 2019
• Affective Forecasting
• Neuroscience

• Find a Branch
• Schwab Brokerage 800-435-4000
• Schwab Password Reset 800-780-2755
• Schwab Bank 888-403-9000
• Schwab Intelligent Portfolios® 855-694-5208
• Schwab Trading Services 888-245-6864
• Workplace Retirement Plans 800-724-7526

... More ways to contact Schwab

  Chat

• Schwab International
• Schwab Advisor Services™
• Schwab Intelligent Portfolios®
• Schwab Alliance
• Schwab Charitable™
• Retirement Plan Center
• Equity Awards Center®
• Learning Quest® 529
• Mortgage & HELOC
• Charles Schwab Investment Management (CSIM)
• Portfolio Management Services
• Open an Account

How to Determine Your Risk Tolerance Level

When you hear "risk" related to your finances, how does it make you feel? Do you see an opportunity for great returns? Do you imagine the "thrill" of investing? Do you become worried that you'll be left with nothing? Do you believe that risk is just an essential part of the investing process?

To better understand your risk tolerance, ask yourself questions like these and think about your behavioral tendencies—such as what actions you'd likely take after experiencing a significant investment loss or what decisions you've made in the past when the markets took a turn for the worse.

Providing an honest answer to these type of questions—and then taking on a commensurate level of investment risk—could help you build a portfolio that you'll stick with, even when market activity makes you nervous.

What is risk tolerance?

Simply put, risk tolerance is the level of risk an investor is willing to take. But being able to accurately gauge your appetite for risk can be tricky. Risk can mean opportunity, excitement or a shot at big gains—a "you have to be in it to win it" mindset. But risk is also about tolerating the potential for losses, the ability to withstand market swings and the inability to predict what's ahead.

In fact, behavioral scientists say "loss aversion"—essentially, that the fear of loss can play a bigger role in decision-making than the anticipation of gains—can color your approach to risk. Since risk tolerance is determined by your comfort level with uncertainty, you may not become aware of your appetite for risk until faced with a potential loss.

Risk tolerance vs. risk capacity

Though similar in name, your risk capacity and risk tolerance are generally independent of each other.

Your risk capacity, or how much investment risk you are able to take on, is determined by your individual financial situation. Unlike risk tolerance, which might not change over the course of your life, risk capacity is more flexible and changes depending on your personal and financial goals—and your timeline for achieving them.

If you have a mortgage, your own business, kids approaching college or elderly parents who depend on you financially, you may be less likely to comfortably ride out a bear market (given your income needs) than if you're single and not holding any major financial obligations.

A financial shock—like job loss, an accident that comes with expensive medical bills or even a windfall—can also affect your investment decisions by altering the amount of risk you're able to afford.

Keeping in line with your goals

When determining your risk tolerance, it's also important to understand your goals so you don't make a costly mistake. Your time horizon, or when you plan to withdraw the money you've invested, can greatly influence your approach to risk.

Your time horizon depends on what you're saving for, when you expect to begin withdrawing the money and how long you need that money to last. Goals like saving for college or retirement have longer time horizons than saving for a vacation or a down payment on a house. In general, the longer your time horizon, the more risk you can assume because you have more time to recover from a loss. As you near your goal, you may want to reduce your risk and focus more on preserving what you have—rather than risking major losses at the worst possible time.

One way to fine-tune your strategy is by dividing your investments into buckets, each with a separate goal. For example, a bucket created strictly for growth and income can be invested more aggressively than one that is set aside as an emergency fund.

Translating risk tolerance into an investment strategy

Completing the Schwab Intelligent Portfolios ® Investor Profile Questionnaire can help you assess your individual risk tolerance. Here, honesty is definitely the best policy—you want the asset allocation mix in your recommended portfolio to most accurately reflect your true tolerance for risk.

Once you know where you fall along the risk spectrum, the next step is to become familiar with typical performance data for your portfolio. The more you know about what you can expect, the smaller the chance that you will react emotionally when times get tough.

Smart investors consider both risk and return. Investments with higher expected returns (and higher volatility), like stocks, tend to be riskier than a more conservative portfolio that is made up of less volatile investments, like bonds and cash. However, even the most conservative portfolio can experience short-term losses due to ever-changing market conditions. This is why it's important to have a diversified portfolio that includes a wide variety of investment options.

Let's say that at the beginning of 1970 you decided to invest $10,000 into one of the three hypothetical asset-allocation models show below. And every year until the end of 2016, you rebalanced your portfolio to make sure you still had the right percentage of stocks, bond and cash. The most aggressive portfolio would have climbed to $892,028, the moderate portfolio would have been valued at $676,126 and the most conservative portfolio would have been worth $389,519.

Morningstar Direct Calculated using daily returns from January 1, 1970 – December 31, 2016.

Indexes used include: Stocks, S&P 500 Index; Bonds, IA SBBI US IT Govt; Cash, IA SBBI US 30 Day TBill. The example is hypothetical and provided for illustrative purposes only. It is not intended to represent a specific investment product. Dividends and interest are assumed to have been reinvested, and excludes taxes and fees. If fees and taxes had been considered, performance would have been substantially lower. Indices are unmanaged, do not incur fees and expenses and cannot be invested in directly. Past performance is no guarantee of future results.

Such wide-ranging results show how taking on extra investment risk can potentially provide a bigger payoff than playing it safe. But more aggressive choices will also put your risk tolerance to the test. Consider what happened to each of these portfolios when times got tough:

• The aggressive portfolio had the highest annualized return but was also twice as volatile as the conservative portfolio. It also skidded more than 44% in its biggest decline during the period.
• The biggest annual decline for the moderate portfolio, with a slightly less volatile mix of assets, was approximately 32%.
• The most conservative portfolio dipped just 14% in its largest decline. However, it also achieved the lowest annualized return over the period.

The closer you get to when you want to access your money, the more those potential setbacks can sting. If you had taken your money out of one of the riskier portfolios after it tumbled—either because you needed the cash, you "followed the herd" of sellers who drove prices down or you simply couldn't stand stomach the pain of losing so much—your returns during that 47-year period would have suffered significantly.

By contrast, when you accurately gauge your limits for investment risk, and then invest in a portfolio that reflects your risk tolerance, time horizon and personal circumstances, you're one step closer to achieving your financial goals.

Reviewing your risk profile is easy with Schwab Intelligent Portfolios. Simply log into your account, click "update profile" to re-launch the questionnaire and update your answers for a different profile.

Diversification and rebalancing strategies do not ensure a profit and do not protect against losses in declining markets.

Please read the Schwab Intelligent Portfolios Solutions™ disclosure brochures for important information, pricing, and disclosures related to the Schwab Intelligent Portfolios and Schwab Intelligent Portfolios Premium programs.

• Search Search Please fill out this field.

What Is Risk Tolerance?

Understanding risk tolerance, aggressive risk tolerance, moderate risk tolerance, conservative risk tolerance.

• Risk Tolerance FAQs
• Financial Advisor
• Portfolio Construction

What Is Risk Tolerance, and Why Does It Matter?

Risk tolerance is the degree of risk that an investor is willing to endure given the volatility in the value of an investment. An important component in investing , risk tolerance often determines the type and amount of investments that an individual chooses.

Greater risk tolerance is often synonymous with investment in stocks, equity funds, and exchange-traded funds (ETFs), while lower risk tolerance is often associated with the purchase of bonds, bond funds, and income funds .

Key Takeaways

• Risk tolerance is a measure of the degree of loss an investor is willing to endure within their portfolio.
• Stock volatility, market swings, economic or political events, and regulatory, or interest rate changes affect an investor's tolerance for risk.
• Age, investment goals, and income contribute to an investor's risk tolerance.
• An aggressive investor commonly has a higher risk tolerance and is willing to risk more money for the possibility of better, yet unknown, returns.
• A conservative investor commonly has a lower risk tolerance and seeks investments with guaranteed returns.

All investments involve some degree of risk and knowing their risk tolerance level helps investors plan their entire portfolio , determining how they invest. Based on how much risk they can tolerate, investors are classified as aggressive, moderate, and conservative.

Risk tolerance assessments are available online, including risk-related surveys or questionnaires. An investor may also want to review historical returns for different asset classes to determine the volatility of the various financial instruments.

One factor that affects risk tolerance includes the time horizon for an investor. Having a financial goal with a long time horizon, an investor may have greater returns by carefully investing in higher-risk assets, such as stocks. Conversely, lower-risk cash investments may be appropriate for short-term financial goals.

An investor's future earning capacity, and the presence of other assets such as a home, pension, Social Security , or an inheritance affect risk tolerance. An investor can take greater risk with investable assets when they have other, more stable sources of funds available. Additionally, investors with a larger portfolio may be more tolerant to risk, as the percentage of loss is much less in a larger portfolio when compared to a smaller portfolio.

Order your copy of Investopedia's What To Do With $10,000 magazine for more wealth-building advice.

An aggressive investor , or one with a high-risk tolerance, is willing to risk losing money to get potentially better results. Aggressive investors tend to be market-savvy with an understanding of the volatility of securities and follow strategies for achieving higher than average returns.

Their investments emphasize  capital appreciation  rather than income or preserving their principal investment. This investor's asset allocation commonly includes stocks and little or no allocation to bonds or cash.

Moderate investors want to grow their money without losing too much. Their goal is to weigh opportunities and risks and this investor's approach is sometimes described as a “balanced” strategy.

Commonly, moderate investors develop a portfolio that includes a mixture of stocks and bonds, perhaps as a 50/50 or 60/40 structure.

Conservative investors are willing to accept little to no volatility in their investment portfolios. Retirees or those close to retirement age are often included in this category as they may be unwilling to risk a loss to their principal investment and have a short-term investment strategy.

A conservative investor targets vehicles that are guaranteed and highly liquid. Risk-averse individuals commonly opt for bank certificates of deposit (CDs), money markets, or U.S. Treasuries for income and preservation of capital.

What Is an Example of a 60/40 Portfolio Structure?

A moderate risk-tolerant investor may choose to invest in a 60/40 structure which may include a 60% investment in stocks, 30% in bonds, and 10% in cash.

What Financial Instruments are Considered High Risk Investments?

High-risk investments include investing in options , initial public offerings (IPO), and foreign emerging markets .

How Does Risk Tolerance Compare to Risk Capacity?

While risk tolerance measures an investor's willingness to take risk, an investor's risk capacity measures their financial ability to take a risk.

U.S. Securities and Exchange Commission. " Assessing Your Risk Tolerance ."

Charles Schwab. " How to Determine Your Risk Tolerance Level ."

• Terms of Service
• Editorial Policy
• Privacy Policy
• Your Privacy Choices

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List

Peer review: Risk and risk tolerance

Stephen a. gallo.

1 Scientific Peer Advisory and Review Services Division, American Institute of Biological Sciences, Herndon, Virginia, United States of America

Karen B. Schmaling

2 Department of Psychology, Washington State University, Vancouver, Washington, United States of America

Associated Data

An anonymized version of these data is available on the Open Science Framework (DOI 10.17605/OSF.IO/FU83D ).

Peer review, commonly used in grant funding decisions, relies on scientists’ ability to evaluate research proposals’ quality. Such judgments are sometimes beyond reviewers’ discriminatory power and could lead to a reliance on subjective biases, including preferences for lower risk, incremental projects. However, peer reviewers’ risk tolerance has not been well studied. We conducted a cross-sectional experiment of peer reviewers’ evaluations of mock primary reviewers’ comments in which the level and sources of risks and weaknesses were manipulated. Here we show that proposal risks more strongly predicted reviewers’ scores than proposal strengths based on mock proposal evaluations. Risk tolerance was not predictive of scores but reviewer scoring leniency was predictive of overall and criteria scores. The evaluation of risks dominates reviewers’ evaluation of research proposals and is a source of inter-reviewer variability. These results suggest that reviewer scoring variability may be attributed to the interpretation of proposal risks, and could benefit from intervention to improve the reliability of reviews. Additionally, the valuation of risk drives proposal evaluations and may reduce the chances that risky, but highly impactful science, is supported.

Introduction

Research funding agencies, including the US National Institutes for Health (NIH), rely on peer review to help identify the most scientifically meritorious and impactful research proposals [ 1 ]. Of concern is evidence that reviewers can discriminate good from bad proposals more reliably then good from great proposals [ 2 – 4 ]. Reviewers are often expected to evaluate beyond their discriminatory power, resulting in relying on biases and preferences [ 5 , 6 ]. Previous studies document cases of low inter-reviewer reliability and more variability between reviewers than between proposals [ 7 – 10 ].

However, little research has examined associations between reviewer characteristics and score variability [ 9 – 12 ]. A potentially important characteristic is risk tolerance, which may reflect a general personality factor [ 13 – 16 ]. There is evidence that reviewers evaluate risky proposals negatively, where the perceived weaknesses are not fully offset by proposal strengths [ 11 , 17 , 18 ]. Examples of such potential risks in research proposals include new or poorly constructed design or methodology, or principal investigators (PIs) transitioning to new areas. It remains unexplored whether reviewers favor differing levels of risk or sources of weakness, although it is clear that the NIH proposal evaluation criteria–approach, significance, innovation, PI, and environment–are not equally predictive of overall scores [ 19 , 20 ].

In this study, we conduct a cross-sectional experiment of 605 NIH peer reviewers’ evaluations of mock primary reviewers’ comments (overall impact statements (OISs)) in which the level and sources of risks were manipulated. Manipulated sources of risk were the PI or the approach and manipulated levels of risk were low or moderate, resulting in four OISs: low risk PI and low risk approach (control OIS); low risk PI and moderate risk approach (approach risk); moderate risk PI and low risk approach (PI risk); and moderate risk PI and moderate risk approach (PI-approach risk). We evaluated the association of proposal risks and reviewer characteristics, including risk tolerance (see Methods ), with reviewers’ criteria and overall scores.

This study recruited reviewers experienced in NIH and NIH-style grant evaluation to participate in an on-line experiment and survey. The survey involved answering self-report measures of demographics; grant review experience, expertise, and evaluative bias; and a personality measure of risk tolerance. The experiment involved rating two basic science NIH-style overall impact statements (OISs) in random order. Participants all rated a positive OIS that had no weaknesses in any criteria category (significance, innovation, approach, investigator, environment) and a second OIS. The second OIS was randomly determined to reflect one of three combinations of weaknesses: weaknesses in the approach only; weaknesses in the PI only; weaknesses in both the approach and PI.

Participants

Participants were recruited from two sources ( Fig 1 ). The first source was comprised of 10,990 people in the American Institute of Biological Sciences’ (AIBS) database who had putatively reviewed for or submitted to one of the funders for whom AIBS had conducted an NIH-style peer review process. Email addresses were extracted from the database: web searches for the current email address were conducted for individuals with multiple email addresses. The second source was comprised of 1678 scientists listed on study section rosters affiliated with the following 6 (1 st wave) + 9 (2 nd wave) 2020 NIH integrated review groups (IRGs): 1 st wave; Cardiovascular and Respiratory Sciences; Immunology; Infectious Diseases and Microbiology; Molecular, Cellular, and Developmental Neuroscience; Oncology Basic Translational; and Vascular and Hematology; 2 nd wave; Biological Chemistry; Cell Biology; Digestive, Kidney, Urological Systems; Endocrinology, Metabolism, Nutrition and Reproduction Science; Genes, Genomes and Genetics; Infectious Disease; Musculoskeletal Research; Oncology Translational Clinical; and Surgical Science, Biomedical Imaging and Bioengineering. These IRGs were chosen based on size and scientific topic area; particularly, we were looking for scientists who would be reasonably familiar with the type of in vivo studies as those described in the OISs. Email addresses were identified by web searches. Only scientists affiliated with US-based institutions with US-based emails were used.

Recruitment flowchart detailing sample gathering from two initial sources, as well as internet searches of bounced emails. Scientists were sent the survey via email (and a reminder email) in 2 waves (September/October of 2020 and February of 2021) and responses were finalized in spring 2021.

From the first source (AIBS), an initial invitation was successfully emailed to 10,594 unique individuals in three groups in September and October of 2020. There were 1430 bounced emails. The 10594 number of successful emails includes emails that were successfully emailed the first time, plus alternate email addresses found by internet research of those with bounced emails ( Fig 1 ). The internet research of bounced emails also revealed people on the original list who were not pursued further if they were determined to be institutional officials, unlikely to be biomedical researchers and have conducted NIH-style peer review (e.g., emails associated with state fish and wildlife divisions), or deceased. This invitation stated how they were identified and invited them to participate if they had done one or more NIH-style reviews in the last 5 years. One reminder invitation was sent two weeks after the initial invitation to those who had not completed the survey nor opted out from further contact. Of those successfully emailed from this first source, 4.5% completed the full survey (n = 475).

From the second source (NIH), an initial invitation was emailed to the 734 unique individuals in one group in early February of 2021 with 13 bounced emails. One reminder invitation was sent two weeks after the initial invitation to those who had not opted out from further contact nor completed the survey.

Another wave was sent to 954 NIH IRG members in late February with 16 bounced emails. These numbers included those found by further email research of those with bounced addresses. From the second source, a total of 7.7% of those successfully emailed completed the full survey (n = 130).

The measures were set up as a Qualtrics TM survey in the following order, unless randomly determined as noted below. First was the consent form, which included a statement of the inclusion criteria, namely having conducted one or more NIH-style reviews in the last 5 years. Demographic information was collected next and included age, academic degrees, gender identity, English as the first language, racial/ethnic identities, and the number of study sections/reviews committees on which the participant had served in the past five years for each of several US research funding agencies (NIH, NSF, AHRQ, DoD, or other). If the participant reported serving on zero NIH, AHRQ, or DoD study sections, which all use similar review criteria and scoring procedures, they were disqualified.

OISs were rated next (see below), followed by self-reported measures of the similarity of their research relative to the OISs (on a scale of 1 to 7 where 1 is extremely similar and 7 is extremely dissimilar) and evaluative disposition (on a scale of 1 to 7 where 1 is lenient/generous and 7 is firm/rigorous). Finally, as a proxy for reviewer risk tolerance [ 13 – 15 ], participants completed the 13-item Openness to Experience scale from the NEO-Five Factor Inventory 3 [ 13 – 15 , 21 ], which has previously established reliability, validity, and norms. The raw scores from the Openness to Experience scale were converted to normed T scores by subject gender [ 21 ], and the T scores were retained for analysis. The T score values of our participants ranged from 28 to 60 with a median of 46 (IQR of 9). Average scores were very similar across all groups of participants. Participants were given an opportunity to add comments about the study in an open text field. Finally, they were asked to provide their email address for the honoraria ($20 Amazon gift card).

Overall impact statements

The four OISs were designed as typical primary reviewer OISs of biomedical research proposals submitted to the NIH that address a specific disease or disability. Proposed work in all of the OISs was at the in vivo level of translatability, but was represented as addressing a critical clinical need and having high potential impact to translate to a clinical setting. The descriptions of innovation and environment did not differ between the OISs, and were written to suggest a significant degree of novelty in the hypotheses, and to have an excellent supporting environment, respectively. The level of risk in the scientific approach and/or with the PI were manipulated in three of the OISs as described below. The PI’s gender was also manipulated, but these data are not examined in this manuscript. The impact statements were constructed to address each criterion–significance, innovation, approach, investigator, environment—and to have 300–400 total words per statement.

Participants received the following instructions: “You are an unassigned reviewer who hears the following summary from an assigned reviewed of an NIH R01 application. In this scenario, you do not have the ability to reference the original proposal or discuss the application with the assigned or other reviewers. Please score the overall impact and also significance, innovation, investigator, approach and environment. You may consult the score guidance table below.” Participants rated each OIS for overall merit and five criteria on the NIH scale of 1 to 9 using whole numbers (1 = exceptional and 9 = poor).

The control OIS was created to reflect an exceptional or outstanding proposal with no discernable weaknesses, i.e., low-risk approach and a low-risk PI. All participants rated the control OIS. Participants also rated one of three manipulated OISs. The three other OISs varied in the PI and/or approach risk: low approach risk and moderate PI risk; moderate approach risk and low PI risk; or moderate approach risk and moderate PI risk. For these 3 OISs, the text was held constant for the topic being investigated, the significance and innovation of the work, and the quality of the environment. The four OISs are included in the S1 File .

We pilot tested these OISs prior to finalizing the text of the statements. Because past research has indicated that reviewers do not discriminate good from great proposals easily [ 3 ], the control statement was intended to reflect a great proposal (score of 1–2) and the manipulated statement was intended to reflect a good proposal (score of 3–5). We tested draft statements with 25 experienced peer reviewers. The pilot data suggested that our manipulations of risk were successful: overall scores for the control statement were significantly better than those of manipulated statements. The pilot subjects were also interviewed and based on their feedback, the OISs were modified slightly. The pilot study and the study reported here were reviewed by the Washington State University Office of Research Assurances (Assurance# FWA00002946) and granted exemptions from IRB review consistent with 45 CFR 46.104(d)(2).

Data manipulation

Data were exported out of Qualtrics TM into comma-separated format to Excel TM . Some variables were re-coded for the regression analyses. Most demographic data variables that were categorical in nature were coded as 0 or 1 (except for gender). These variables included race (coded as 0  =  White, 1  =  non-White), gender (1 = female, 2  =  male, 3 = non-binary), English as a first language (1 = yes, 0 = no), PhD degree (1 = yes, 0 = no), and MD degree (1 = yes, 0 = no). The measure of risk tolerance (NEO Openness to Experience) was z-score transformed, because of their large range.

Statistical analysis

Data were imported from Excel to the statistical program R ( https://www.r-project.org ), and analyses were performed using the ordinal, rcompanion, and ggplot2 packages. Median values and distributions of overall and criteria scores were examined for the four OISs using violin plots. In some cases, non-parametric tests (e.g. Mann-Whitney or Kruskal–Wallis tests) were conducted to confirm differences between risk scenarios using StatPlus TM software. To examine how well participants could detect risk, receiver operating characteristic (ROC) curves were calculated for the overall score using a series of thresholds over the full scoring range (1–9). We defined a true positive as a participant identifying correctly that a scenario is risky by scoring worse that the threshold. We defined a false positive as a participant identifying incorrectly that the control scenario is risky. For each risk group, the area under the curve (AUC) and the corresponding 95% confidence intervals were calculated.

Given the ordinal values and the cross-sectional, nested nature of the OIS scoring data (two sets of scores nested within each participant), multi-level ordinal regression models were used to examine the relationship of the predictors to the dependent variables of criteria and overall scores. Specifically, cumulative link mixed models were used, fitted with the Laplace approximation and variance, which looked at fixed effects from predictor variables and random effects from participants (two sets of scores per participant). An intra-class coefficient (ICC) was derived from the variance estimate of the final regression model to assess the inter-rater consistency across all participants [ 22 ]. We also estimated Cronbach’s alpha to measure the within-rater consistency between the rating of control and risk scenarios for overall and criteria scoring.

Starting with an intercept-only baseline model, predictor variables were added in blocks and models were compared via log likelihood for successive improvements in fit. Nagelkerke-R 2 values were calculated as well, to compare models to baseline. Interaction effects were examined as variable blocks were added to the model, and if none were observed, were not included in the proceeding models. These models were applied to the overall and individual criteria scores for each OIS risk scenario. Relative likelihood profile functions were calculated using adaptive Gauss-Hermite quadrature approximation (with 10 quadrature points) for the random effects with 95% confidence intervals based on the profile function. For model comparisons, participants with missing data were removed for consistency.

Post-hoc visualizations included LOESS linear fits of the overall score for each criterion broken down by risk scenario. Jittered scatterplots using local regression (LOESS) fitting and boxplots were also used to examine relationships between variables. Also explored were the effects of the order of OIS presentation (e.g., control first) and whether the participants were recruited from the AIBS database or from NIH rosters. Some fixed effects ordinal regression models were applied to specific risk scenarios to further analyze the data.

Response completion and demographics

In total, 592 (AIBS) + 153 (NIH IRG list) individuals completed some of the survey. Of these 745 individuals, 140 individuals did not complete the survey; 16 stopped at the consent form; 25 at the demographic section; 99 at the OISs. Of this group of participants with incomplete data, 29% (41) did not provide demographic data. Of the 99 that provided demographic information, 30% were female, 78% were White, 76% were native English speakers, 74% had PhDs, and 68% were 20–40 years since their degree; all of these proportions were similar to the demographics of the participants with complete data in the main sample ( Table 1 ). However, only 40% of participants with incomplete data had been part of 10 or more review panels in the last three years (compared to 59% from the main sample; Table 1 ) and only 70% had taken part in an NIH panel (compared to 95% in the main sample; Table 1 ). Thus, many of those who did not complete the survey did so due to limited review experience. Their exclusion is consistent with review experience as the major inclusion criteria for our study.

Total N is 605, PI risk group (199), Approach risk group (204), PI and Approach risk group (202).

Among the 605 participants (475 from the AIBS sample, 130 from the NIH sample) who had completed the survey, 7.6% (n = 46) were missing data as follows: risk tolerance, 4; racial identity, 3; degree, 1; degree year, 16; research similarity, 24; evaluative predisposition, 8.A total of 559 participants had no missing data at all (460 from the AIBS sample and 99 from the NIH sample). In the final regression models that utilized a maximum likelihood function, the full data set (N = 605) was used, but the model comparisons utilized the reduced data sets (N = 559).

Median survey duration was 13 minutes (IQR = 11 minutes). Participant demographics are listed in Table 1 ; most participants are White male PhDs with English as their first language. The majority of participants received their most recent degree 20 to 40 years ago. Just over 95% of participants had reviewed for NIH in the last 3 years, with the majority participating in up to 20 review panels over that 3-year period. Demographics were found to be relatively consistent across participants rating the three risk scenarios ( Table 1 ). These demographics are very similar to those reported by NIH of their reviewers. [ 23 ] An anonymized version of these data is available on the Open Science Framework (DOI 10.17605/OSF.IO/FU83D).

Multi-level ordinal regression modeling

We examined variance inflation factors (VIFs) for all variables used in the model and found no case where VIF>10 ( S1 Table ), which we used as a criterion for inclusion in the regression [ 24 ]. From baseline, we added risk (a contrast variable examining either PI-only, approach-only, or PI-approach scenarios compared to the control) as a predictive factor to the model for overall score, and found a large improvement in fit compared to baseline ( Table 2 ) . Demographic variables were then added as a block, then the rating of their research similarity relative to the OIS and evaluative predisposition, none of which improved the model. We then added the measure for risk tolerance (NEO Openness to Experience scale), to the model, but with no improvements in fit or variance explained. We then added the five criteria scores (significance, innovation, investigator, approach and environment) as a block to the model, and found a significant improvement in fit when the model was compared to baseline. Similar steps were taken for all criteria scoring models.

* p< 0.05

** p<0.01; 95% CI in parentheses; each successive model is compared to previous via -2LL (a fixed intercept model was used as baseline); Nagelkerke R 2 was calculated comparing to baseline model.

Manipulation check

The experimental manipulation was effective. Probability densities, distributions (via overlayed boxplots) and median values (in red) of the overall and criteria scores for each OIS are shown in violin plots ( Fig 2 ). Lower scores are more positive, i.e., 1 = exceptional and 9 = poor. The three OISs with manipulated risk, intended to represent “good” proposals, were rated consistent with this descriptor, with less favorable scores and more dispersed scoring distributions than the control OIS, which was intended to represent a “great” proposal, and was rated in the outstanding range ( Fig 2 and S2 Table ). The median values of the overall score are similar for the PI risk and approach risk OISs, but the scores for the approach risk OIS are more dispersed and positively skewed (U = 14,113, n1 = 205, n2 = 199, p < 0.001; effect size = 0.55). The overall score of the PI-approach risk OIS was rated more mediocrely than either PI or approach risk OISs, and the scoring distribution and probability densities are dispersed. This dispersion is reflected in the ICC measurement, which across all participant groups was 0.15, reflecting the significant spread of scores within the three risk groups. While the magnitude differed across reviewers, the majority of reviewers agreed in each risk case (65%, 79% and 88% for PI, approach and PI-approach risk groups, respectively) that the manipulated OIS should receive a worse score than the control OIS. Based on the ROC plot ( Fig 3 ), participants were observed to have excellent discrimination in determining the presence of risk; participants in the PI risk group had the lowest AUC (0.74 [0.70, 0.79]), with the approach risk group having an increased AUC (0.83 [0.79, 0.87]) and the PI and approach risk group having the highest AUC (0.91 [0.88, 0.94]).

Median values (in red) and distributions of overall and criteria scores for different risk scenarios (violin plot; N = 605). A. Overall; B. Significance; C. Innovation; D. Investigator; E. Approach; F. Environment.

Receiver operating characteristic plots were calculated for the overall score using a series of thresholds over the full scoring range (1–9). Proportions of true positives versus false positives were plotted for each risk group.

Criteria scores also generally reflected the risk manipulations ( Fig 2 and S2 Table ). Significance and innovation scores had constant median values across OISs. However, the score distributions of unmanipulated criteria were affected by the manipulation of PI or approach: e.g., environment was evaluated more poorly in manipulated OISs as compared to control. Also, the PI was evaluated more poorly in the approach risk OIS than control, despite no manipulation of PI in this OIS. Moreover, approach was evaluated more poorly in the PI-approach risk OIS than in the approach risk OIS.

The risk manipulations resulted in greater dispersions in overall and criteria scores. To further examine the relationship between score dispersion and risk, mean scores were plotted against scores’ standard deviation for overall and criteria scores. There were strong associations, whereby riskier scenarios resulted in more dispersed scores ( Fig 4 ).

Plots of the standard deviation of scores (overall and criteria) versus mean scores for different OIS risks. A. Overall; B. Significance; C. Innovation; D. Investigator; E. Approach; F. Environment.

Multi-level ordinal regression modeling—overall score

We created a series of multi-level ordinal regression models to examine OIS risk, criteria scores and their interactions, demographic variables and risk tolerance as predictors of overall scores. The intercept of our baseline model was allowed to vary across participants; we then added blocks of variables as fixed effects and compared successive models for improvement in fit using log likelihood. Random slopes were also explored but were not found to improve model fit. To facilitate model comparisons, only participants without missing data were used (559 participants), but the final model (which utilized a maximum likelihood function and can handle missing data) used the complete data set (605 participants). The final model included interactions between the five criteria variables and risk scenario variable, with a Nagelkerke R 2 = 0.78 ( Table 2 ) . Random effects of participants accounted for a significant amount of the variance in our final model. The contributions of individual predictors are listed in Table 3 : OIS risk, significance, innovation and approach scores, and interactions of all the criteria scores except for environment with OIS risk were significant predictors of overall scores ( Table 3 ). Risk tolerance did not predict overall scores.

** p<0.01

*** p<0.001

Multi-level ordinal regression modeling—criteria scores

A series of multi-level ordinal regression models was used to examine predictors of criteria scores, using a fixed-intercept baseline with random effects for participants, successively adding blocks of variables and comparing model fit. These results are presented in S3 – S12 Tables . For all criteria score models, OIS risk was observed to significantly predict scores. Risk tolerance did not improve model fit for any criteria score other than approach for which–interacting with OIS risk—it was a statistically significant but weak predictor (see jittered scatterplot with LOESS linear fits of approach scores versus risk tolerance for different OIS risk in S1 Fig ). Final models ranged in their fit compared to baseline models: the final innovation score model had a Nagelkerke R 2 of 0.04 and the final PI score model had a Nagelkerke R 2 of 0.57.

Overall and criteria score relationships

LOESS linear fits were used to visualize the relationships between overall and criteria scores for the different risk OISs in a series of jittered scatterplots ( Fig 5 ). Overall scores for the risky OISs were less sensitive to criteria scores than was the control OIS, as is evidenced by the steeper slopes for the control OIS. The plots of overall score with approach are exceptions, where the slopes for all OIS risks are similar. The strength of these dependencies can be visualized via the coefficients of criteria predictors to overall score in simple ordinal regressions by OIS risk; Fig 6 shows that significance and innovation criteria scores are stronger predictors of overall scores in the control OIS than in the risky OISs, while approach is a strong predictor of overall score for all of the OIS conditions.

Scatter plots and LOESS fits of overall score versus criteria scores for different OIS risks (Linear Fits). A. Overall vs Significance; B. Overall vs Innovation; C. Overall vs Investigator; D. Overall vs Approach; E. Overall vs Environment.

Comparison of fit coefficients from four ordinal regression models of overall versus criteria scores. Four regression models (one for each OIS) were compared for relative importance of the 5 criteria scores in determining the overall score.

Control versus manipulated OIS scores

Participant random effects are significant predictors of overall scores: one contributing factor may be the contrast of the manipulated and control OISs within participants; an examination of this relationship for different risk OISs is shown in Fig 7 . These graphs indicate that manipulated and control scores are associated for overall and criteria scores, and suggest that, although these relationships change between risk scenarios, there is significant intra-rater consistency. However, across all the data, our measurements of Cronbach alpha of within-rater consistency between control and manipulated scores indicated the ratings of the manipulated components (the PI and approach) and the overall score are less consistent within raters than are the unmanipulated components (significance, innovation, environment; S13 Table ).

Scatterplots and LOESS linear fits of scores from three manipulated OISs with one or more sources of risk compared with the control score for both the overall and all criteria scores. A. Overall; B. Significance; C. Innovation; D. Investigator; E. Approach; F. Environment.

Risks in proposal evaluation

Risk was observed to be a strong predictor of overall and criteria scores, with which it interacted significantly. Risk impacted the scores for criteria that were not manipulated and were written as proposal strengths and assets (e.g., significance and innovation). As risk increased, the predictive power of these proposal assets towards the overall score diminished. In contrast, the approach criteria score was observed to strongly affect participants’ overall evaluations. These results suggest that reviewer evaluation of proposals’ risks dominates over that of proposal assets in OISs with proposal risk.

These results have broad implications for grant review and are aligned with previous studies that proposal scores are associated with the number of weaknesses in reviewers’ critiques but not with strengths [ 25 ]. Our findings suggest that evaluative rewards from innovative and potentially significant ideas are drowned out by evaluative penalties from risky aspects of proposals. These differences between rewards and risks may be partly due to the level of information inherent in these aspects of a proposal. For example, the evaluation of the longer-term benefits of research can be difficult. The OISs stated, “this work has high potential for translational impact,” but the specific impact was not described. More specific information was provided about risk and its impact on project success: “if the investigators find substantial toxicity or side effects in Aim 1, the other Aims can no longer be completed.” Information availability shapes decision making. The theory of bounded rationality, that people make rational decisions that are bounded by their expertise or knowledge [ 26 , 27 ], is relevant for peer review decisions [ 10 , 11 ]. Low myopic problem representation biases, where differences in information levels can cause an overweighting in comparative judgements, may also be pertinent [ 28 – 30 ]. More research is needed to apply these existing theories to peer review and explore how the clarity of translational impact may influence scoring for risky proposals.

Risk impact

The source of risk (PI and/or approach) was an important factor in overall scores, with participants penalizing approach risks more than PI risks. The approach risk OIS described the impact of the risk to the experiments (i.e., dependency of Aims may prevent completion of the project). However, in the PI risk OIS, the inexperience of the PI was described as possibly affecting project completion, but how this risk would otherwise impact the study was not articulated. Thus, the information about the likelihood and potential impact of PI risk was less than approach risk, and was likely evaluated as relatively minor compared to the specific approach risk information. When both manipulations (PI and approach risk) are present in the same OIS, there may be a perceived increase in information, leading to the perception that these combined risks would impair project success.

Risk intensity and scoring dispersion

As risk increased, the accurate detection of its presence by participants also increased, but so did the dispersion of participants’ scores, suggesting that proposal risk is an important source of inter-reviewer variability for projects of moderate-to-high quality. This relationship between risk and score distributions was found for overall scores and criteria scores. Pre-review orientations typically do not include guidance regarding how to translate risks into final scores. Reviewers may need training regarding risk acceptability and tolerance, and specific risk scenarios that could be evaluated positively, especially in distinguishing good from great proposals.

Risk tolerance

The self-report questionnaire measure of risk tolerance, the NEO openness to experience scale [ 21 ], did not predict overall or most criteria scores, the exception being a very weak but significant predictor of the approach criteria score and its significant interaction with OIS risk. The low level of association between risk tolerance and OIS risk was surprising given the centrality of risk in our OIS manipulations. In the control OIS, risk tolerance was associated with more favorable approach scores, but this relationship largely disappeared in the manipulated OISs. Reviewer evaluation of approach in an outstanding proposal–as with our control OIS–may represent reviewer predisposition to risk. The literature suggests risk tolerance can be self-reported or revealed experimentally, and that the two types of measures can lead to different conclusions [ 31 ]. The approach score of an outstanding proposal could also simply reflect the level of leniency or stringency of participants as evaluators, despite our self-reported measure of leniency (see Methods ) not being correlated with scores. Bias related to reviewers’ stringency has been reported for large studies of NIH review data [ 32 ]. In our study, this leniency effect depends on the riskiness of the OIS, as well as whether the component being scored was a manipulated component in the OIS. The least consistency between control and manipulated OIS scoring was in the approach score, despite the fact that the approach criteria was a key overall score-driving criteria in our and previous studies [ 20 , 19 , 33 ]. Understanding the interaction between evaluative stringency, sources of risk, and review criteria is an area for future investigation.

Potential limitations

A limitation of this study is that evaluating proposals based on reading OISs is not representative of actual grant review, where scientists have access to the whole proposal, hear summaries from multiple assigned reviewers, and discuss the proposal before scoring. However, previous studies have shown that most unassigned reviewers’ scores are highly correlated with the assigned reviewers’ scores, which change minimally after discussion [ 34 ]. Another limitation is the low inter-rater agreement among participants, consistent with previous reports of grant review [ 7 ]; however, for all risk scenarios, the majority of reviewers in our study agree that the manipulated risk scenario should receive a worse score than the control. More research should focus on how reviewers translate this penalty into a numeric score and on developing review procedures which reduce variability due to this translation. Also, our response rates for this experiment were low, which could affect the generalizability of this work; however, these response rates are on par with similar previous surveys on peer review [ 35 – 37 ] Given the demographic similarity of our sample to NIH reviewers [ 23 ] and our inclusion criteria of review experience, this sample is likely to be highly representative of grant reviewers at NIH. Further, participants who started but did not complete the study may be attributable to limited review experience based on the data provided. Finally, our risk tolerance measure was not predictive of scores, and better alternatives may need to be created specific to peer review.

Conclusions

This study found proposal risks–more than proposal strengths—to be a score-driving factor in the peer review of moderate- to high quality proposals. These results are of concern insofar as evaluations of highly innovative but risky proposals may be attenuated. Inter-reviewer score variability may depend on the perceived likelihood and impact of risks on project success and may represent an area of needed intervention for peer reviewers. More conversations in the scientific community about values and priorities in peer review would be beneficial, e.g., of acceptable levels of risk and potential failure in grant proposals, and how to appropriately balance the assessment of risks with assets. Ultimately, expectations of peer reviewers should be tempered with research on peer review processes and limitations.

Supporting information

Risk Tolerance (NEO Openness Score) vs Approach Scores (LOESS plots).

VIF analysis for mixed ordinal regression of overall scoring data.

Kruskal-Wallis rank tests examining differences in participant scoring over the four risk scenarios: the control and the 3 manipulated OISs.

Significance Score–Multi-level ordinal regression models made with the reduced data set for direct comparison (n = 559).

Cumulative Link Mixed Model of Significance Score fitted with the Laplace approximation from the total data set (605 participants).

Innovation Score–Multi-level Ordinal Regression models made with the reduced data set for direct comparison (n = 559).

Cumulative Link Mixed Model of Innovation Score fitted with the Laplace approximation from the total data set (605 participants).

Investigator Score–Multi-level Ordinal Regression models made with the reduced data set for direct comparison (n = 559).

Cumulative Link Mixed Model of Investigator Score fitted with the Laplace approximation from the total data set (605 participants).

Approach Score–Multi-level Ordinal Regression models made with the reduced data set for direct comparison (n = 559).

Cumulative Link Mixed Model of Approach Score fitted with the Laplace approximation from the total data set (605 participants).

Environment Score–Multi-level Ordinal Regression models made with the reduced data set for direct comparison (n = 559).

Cumulative Link Mixed Model of Environment Score fitted with the Laplace approximation from the total data set (605 participants).

Cronbach’s alpha and correlation of within-rater consistency between the rating of control and risk scenarios.

Text of the four OIS impact statements.

Acknowledgments

The authors appreciate the assistance of the following WSU students in finding email addresses of potential participants: Anne Blake-Nickels, Kathleen Brandt, Hailey Landon, Ashley Marshall, Tiffany Nguyen, Molly Permin, and Haylee Stack.

Funding Statement

This work has been supported by the Social, Behavioral and Economic Sciences Office of Multidisciplinary Activities at the National Science Foundation ( https://www.nsf.gov ), under the standard grant award numbers 1951132 (SG) and 1951251 (KS). The funders (NSF) had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability

Risk Tolerance

By: nguyentuegiang   •  Essay  •  1,536 Words  •  May 5, 2011  •  1,872 Views

As mentioned by Garble (2000), risk tolerance is the largest limit of uncertainty which investors can accept when they decided to invest on some kinds of assets. Besides, it is also defined as the way an investor behaves to risk in the planning industry (Hallahan et al, 2003). Harlow and Brown (1990) stressed that estimate the risk tolerance degree is significantly helpful because it presents investors attitude in the exchange of what investor can achieve and how much risk investor have to stand for. Bodies et al (2009) provide three kinds of investors with different risk tolerance score. They are risk adverse investor, risk-neutral investor, risk lovers with the aversion toward risk arranged in the decrease line, respective. It implies that when investors have to make decision with comparable gains, investors will have alternative options depend on their risk aversion. Risk averse investors are not satisfied with facing much risk so they always refuse to invest in portfolio with the risk insurance is zero (Bodies et al, 2009). Although risk averse only achieve minimum gains but they have safety risk level. Besides, to risk loving investors, portfolio with the highest possible gain and the highest acceptable risk attracts them the most (Rutterford and Davison, 2007). Hence, risk lovers are investors who dare to take part in almost risky portfolio with the optimistic attitude. On the other, neutral-risk investor seems to be insensitive with risk level. They involve completely in expected return rate and they usually evaluate the risky capacity through it (Bodies et al, 2009). With each characteristic of each investor type divided by risk tolerance, it is obvious that risk attitude have effect on the making investment decision. There is a fact that when the same portfolio is given, some investors will accept it while others give up. It is because the different risk attitude of each investor. For example, with the portfolio which has zero risk premiums, risk averse will stay away from it, however, risk loving loves it (Bodies et al, 2009). According to Brentani (2004), Utility scores is the method which rely on expected return as well as risk level to arrange various investment portfolio. It is measured by the formula U=E_((r)- ) 1/2 A?^2. In this formula, it has a component is A which represent a specific number of investor's risk aversion and E(r) is the expected rate of return (Bodies et al, 2009). Hence, it is easy to see that there is a mutual relationship between risk tolerance and utility scores. This author (2009) also highlights that with the same risk and return, utility score will different among investors with different risk tolerance.

The initial portfolio of UP investment team

Sharpe (1992) particularly defines the asset allocation as the way how to choose all different assets options to put on an investment portfolio through enormous asset categories. This author also highlights the characteristic of asset allocation is variable. The reason is that it is important to find the most effective mix among various classes of assets. However, based on the market conditions, macro economic situation and other affected elements, each investor will need to evaluate the performance of portfolio under time interval to re-allocate the assets percentiles and assets types. Nuttall (2000) represents the trusted ideas from a study is 93.6% of portfolio return depends on the way investor arrange asset. This stresses the significant importance of assets allocation in investment activities. However, it is also needed to set a policy statement as a guiding star for investment cycle (Reilly and Brown, 2009) which often include investment objective and the strategy to implement.

The author's investment team is called UP team. UP provided a policy statement as follow. Firstly, about the objectives, the goal for investing is to gain enough money for letting the children to college and university. The time investment horizon is 10 years. In addition, Up want to achieve the aggressive growth with the expected growth percentage is 6-8% per annual. Secondly, about the investment strategy, UP applied an active strategy is for constructing the portfolio. Thirdly, UP asset allocation benchmark model was based on Fidelity Aggressive Growth model with 60% for stock, 30% for bond and 10% for cash. Because the risk tolerance score of UP is risk loving, UP accept to gain higher return by taking more risky from choosing high stock proportion. Besides, cash are kept to face with problem when other asset classes don't work effectively.

For stock decision, UP decided to invest on four main markets (UK, US, France, Russia) - eight sectors - fifteen specific companies. The reason for UP to diversify stock funds is too decrease the possible risk UP could face. In order to choose sectors, UP based on the performance chart over the last 6 month of

Assessing risk tolerance

Assessing risk tolerance is a critical step in determining the appropriate trade off between the risk and expected return of a portfolio. Selecting a mix of risky and risk-free assets (or higher-risk and lower-risk assets) is one of the most important decisions in designing a portfolio. Investors must develop a rational assessment of risk tolerance to make rational decisions about portfolio design.

Most approaches to assessing risk tolerance consider these criteria in one way or another:

• Attitude toward risk
• Investment time horizon
• Other factors that affect ability to compensate for investment losses, for example, net worth, stability of income, future liabilities, and flexibility of goals.

This article presents several frameworks to help you to assess your risk tolerance. It uses the term risk tolerance to describe both an investor's capacity to bear risk and their attitude toward risk. [1] However, some authors use the term risk tolerance to describe only an investor's attitude toward risk; that is, the psychological and emotional aspects related to taking risk. To avoid confusion, the article uses "determining appropriate risk exposure" instead of "assessing risk tolerance" for frameworks that use the term risk tolerance to refer only to an investor's attitude toward risk.

• Risk tolerance

Risk is the uncertainty (variation) of an investment's return, which does not distinguish between a loss or a gain. However, investors usually think of risk as the possibility that their investments could lose money.

Risk tolerance is an investor’s emotional and psychological ability to endure investment losses during large market declines without selling or undue worry, such as losing sleep.

To know whether a portfolio is right for your risk tolerance, you need to be brutally honest with yourself as you try to answer the question, "Will I sell during the next bear market?"

Knowing your emotional tolerance for investment risk means knowing yourself and your unique goals and needs - and it is not easy. The next sections show different ways you can assess your risk tolerance.

Boglehead's Guide To Investing

The Bogleheads' Guide To Investing discusses determining appropriate risk exposure in the chapter on asset allocation . The context is designing an efficient portfolio and "staying the course".

The book explores four areas:

• Personal financial situation

Examples of goals are saving for a home, a child's education, or retirement.

Stocks are suitable for long time frames , whereas medium and short-time frames require less risky investments.

Risk tolerance is about the psychological and emotional ability to stick with the investment plan during large market declines. Whether or not you think you could sleep well at night (the sleep test ) with your current asset allocation (proportions of stocks, bonds and cash) is one indicator of risk tolerance.

Your own personal financial situation affects how much risk and what types of risk are appropriate. Stability of income and net worth affect the need to take risk. Investors with higher guaranteed retirement incomes (for example, a pension or social security) or high net worth do not need to take as much risk; that is, they have a lower requirement to invest in stocks.

The book refers to three tools, along with an investor's own experience, to help factor risk into your asset allocation:

• John Bogle , founder of Vanguard, suggests that a rough "rule of thumb" to hold your age in bonds .
• Consider the maximum decline to expect with various stock/bond ratios. The original version of the book presents maximum declines in the 2000 to 2002 bear market, but the maximum declines in the 2008/2009 bear market were worse; a common rule of thumb is to be prepared for a 50% loss in the stock portion of your portfolio.
• Vanguard provides an online questionnaire and suggested asset allocations. [2] The book includes a version of the questionnaire as an Appendix, but see the Vanguard website for the latest version.

The book presents various sample portfolios based on these "stages in life":

• Young investor
• Middle-aged investor
• Investor in early retirement
• Investor in late retirement

Larry Swedroe: ability, willingness, and need to take risk

Bogleheads author Larry Swedroe suggests that investors evaluate their risk tolerance by considering their ability, willingness and need to take risk. [3] [4] [5]

Ability to take risk involves investment time horizon, liquidity needs, stability of earned income, and the flexibility to adapt if the portfolio does not achieve its expected returns (that is, "plan B").

Willingness to take risk is characterized as the eat well/sleep well trade-off. Taking more risk is required to enable the possibility of higher expected returns (eat well). However, if investors take more risk than they are emotionally able to handle, then it is likely that they will abandon their investment plans if their portfolios suffer sufficiently severe losses. So it is unwise for investors to take so much risk that they will be unable to sleep well during the inevitable stock market downturns.

Need to take risk is related to the investment goal. If the goal requires a higher expected return, an investor needs to take more risk with the expectation that doing so will result in a higher return. Perhaps more importantly, once an investor has "won the game" by accumulating sufficient wealth, it is unwise to take more risk than is needed, since the value of additional gains is much less important than the consequences of severe losses.

William Bernstein

Bogleheads author William Bernstein discusses determining appropriate risk exposure in the context of deciding on your allocation between stocks and bonds based on age and risk tolerance (attitude toward risk), at least as a starting point. [6]

A very young investor has many years to take advantage of the inevitable bear markets by buying stocks at depressed prices. At the other extreme, a retired person typically is spending down savings, and therefore an extended bear market is more likely to endanger the likelihood that a stock-heavy investment portfolio will last through retirement.

In finance terms, younger investors have more human capital ; that is, their total future earnings are much larger than their investment portfolio. By contrast, a retired investor who is not working has no human capital. Human capital can be considered as a bond-like investment, so the high value of a young investor's human capital can justify a higher allocation to stocks.

Middle-aged investors fall somewhere in the middle, so perhaps might consider dividing their portfolio evenly between stocks and bonds.

Other than age, risk tolerance is the other major consideration in determining asset allocation. Bernstein uses risk tolerance in the sense of attitude toward risk. Bernstein suggests that investors can evaluate their risk tolerance based on how they reacted to the financial crisis during 2008 and early 2009:

• Sold: low risk tolerance
• Held steady: moderate risk tolerance
• Bought more: high risk tolerance
• Bought more and hoped for further declines: very high risk tolerance

Bernstein mentions age in bonds as the most common age-based rule of thumb, and suggests that investors might modify this based on risk tolerance as follows (stated in terms of percentage point increase or decrease to the stock allocation):

• Very High: +20%
• Moderate: 0%
• Very Low: -20%

For example, a 50-year-old with moderate risk tolerance might hold a 50/50 stock/bond allocation (age in bonds), while a 50-year-old with very high risk tolerance might hold a 70/30 stock/bond portfolio.

Bernstein emphasizes that this is just a starting point. A very wealthy, frugal retiree might have a higher stock allocation since the chance of running out of money is low, and the investment time horizon is relatively long considering the lifetimes of their heirs and other beneficiaries. At the other end of the spectrum, a retiree whose annual living expenses consume a relatively large percentage of retirement savings might consider reducing spending and use most retirement savings to purchase a fixed annuity.

Bernstein points out that investors who have not lived through serious market declines tend to overestimate their risk tolerance. He strongly urges that investors invest conservatively until they have experienced their first bear market. [7] This gives them the chance to determine how they will behave as they see the value of the stock portion of their portfolio decline significantly; for example, will they have the discipline to rebalance from bonds to stocks when it seems most frightening to do so?

Bernstein presents a criterion he describes as the "equipoise point" to help with the stock/bond allocation decision. This is the point at which the pleasure of seeing the stock portion of your portfolio increase in value during a bull market just offsets the regret of not having a higher allocation to stocks. It also is the point at which the pain of stock losses during a bear market counterbalances the positive feelings provided by your bonds (which can be rebalanced into stocks at lower prices).

Rick Ferri: lifecycle model

Bogleheads author Rick Ferri suggests using a modified version of the age in bonds rule of thumb to help decide on the split between higher-risk and lower-risk assets, and presents asset allocation decisions in terms of a life-cycle investing framework. [8]

Age is an important factor in determining asset allocation. Older investors have less time to make up losses, and if still working, have fewer years to replace those losses from savings. However, applying the simple age in bonds rule of thumb is more appropriate for younger investors than for older investors.

The financial situations of younger investors are less diverse, so holding your age in bonds probably is a reasonable guideline. Financial assets are small relative to future earnings, so large stock market fluctuations will have relatively small impact on long-term wealth. A 70/30 stock/bond allocation probably is fine for most 30-year-olds.

However, there is much more diversity in the financial situations of older investors; for example, accumulated wealth, access to pension plans, family size, and so on. Although you can use age in bonds as a starting point, it is much more likely that you will need to make significant adjustments based on your personal financial situation. The concept of "asset allocation age" aims to account for these factors.

For example, investors whose wealth is large relative to spending needs can afford to take more risk, and therefore have asset allocation ages that may be higher than their chronological ages. Examples of factors to consider in determining asset allocation age are income needs, retirement needs, pensions, living expenses, bequeathing goals, passive income sources, attitude toward risk, debt, and possible inheritances.

You can also use a life-cycle investing model to help decide on how to allocate portfolio assets between higher-risk stocks and lower-risk fixed-income investments. There are four life phases to consider:

• Early savers, typically ages 20 to 39, with an appropriate allocation to stocks in the range of 60% to 80%.
• Midlife accumulators, typically ages 40 to 59, with an appropriate allocation to stocks in the range of 50% to 70%.
• Transitional retirees, typically ages 60 to 79, with an appropriate allocation to stocks in the range of 30% to 70%.
• Mature retirees, typically age 80 and up, with an appropriate allocation to stocks in the range of 20% to 60%.

Developing a good understanding of how you are likely to react to market volatility and bear markets is also critical in determining an appropriate asset allocation. Risk tolerance is defined as "a measure of the amount of price volatility and investment loss you can withstand before changing your behavior."

Although risk tolerance questionnaires can be useful as a starting point in discovering your risk tolerance, they generally are not sufficient. An asset allocation stress test is another useful tool. This involves working through a bear market scenario, month by month, using a target portfolio with real dollar amounts, and determining if you are likely to stick with a rebalancing plan as the value of the portfolio declines.

Daniel Solin

Daniel Solin is the author of The Smartest Investment Book You'll Ever Read , listed in the Bogleheads Book recommendations and reviews , and the subject of this Taylor's Gem .

Chapter 29 in another one of Solin's books, The Smartest Money Book You'll Ever Read , is titled Assessing Your Risk Capacity . Solin presents five major factors for assessing risk capacity: [9]

• Time horizon and liquidity needs. Longer time horizon and lower shorter-term liquidity needs increase risk capacity.
• Income and savings rate. Higher income and savings rate increase risk capacity.
• Net worth. Higher net worth increases risk capacity.
• Attitude toward risk. This is what most people refer to as risk tolerance , and is the emotional component of handling losses.
• Investment knowledge. The better your understanding of portfolio theory and the risk-return trade off, the higher your risk capacity.

Solin provides a Risk Capacity Survey on his website. There is a quick survey with five questions, a 25-question "complete" survey, and a 19-question survey for 401(k) plans.

CFA Institute

CFA Institute (Bodie, Kane, Marcus, 2008, chapter 21): Investors and Objectives, Investor constraints (Liquidity, Investment Horizon, Regulations, Tax Considerations, Unique Needs.

Self-assessment questionnaires

Various advisory services offer self-assessment questionnaires to help determine your risk tolerance and define an initial asset allocation.

For example, Vanguard offers the following tool:

• An Investor Questionnaire . By answering a few questions, you can quickly determine an asset allocation. This tool does not recommend any funds.

These tools are useful with caveats. Their real purpose may be to protect the firm by putting the client on record as accepting a certain level of risk. [note 1] [10]

The usefulness of the questionnaire has recently come into question, as it does not address three factors that have the greatest impact on a person’s attitude toward risk: [10]

• Their hereditary preference for taking on financial risks
• The influence of friends and acquaintances on framing their opinions
• Their life experiences, especially in the formative years

However, this kind of tool is a worthwhile exercise because it at least asks a question that is different from saying "What percentage of stocks do you think you are comfortable with?"

• ↑ Regulators often require a firm to determine a client's risk tolerance before that firm can recommend investments.
• Risk and return
• Risk and return: an introduction
• Risk and return: application
• Bogleheads' Guide To Investing
• ↑ "Investor Questionnaire" . Vanguard . Retrieved July 18, 2023 .
• ↑ Larry Swedroe (February 3, 2014). "Asset Allocation Guide: How much risk should you take?" . CBS Moneywatch . Retrieved July 18, 2023 .
• ↑ Larry Swedroe (February 12, 2014). "Asset Allocation Guide: What is your risk tolerance?" . CBS Moneywatch . Retrieved July 18, 2023 .
• ↑ Larry Swedroe (February 19, 2014). "Asset Allocation Guide: How much risk do you need?" . CBS Moneywatch . Retrieved July 18, 2023 .
• ↑ William Bernstein (2010). The Four Pillars of Investing . McGraw Hill. pp. 75–80. ISBN   978-0-07-174705-9 .
• ↑ William Bernstein (2010). The Four Pillars of Investing . McGraw Hill. p. 124. ISBN   978-0-07-174705-9 .
• ↑ Rick Ferri . All About Asset Allocation (2nd ed.). McGraw Hill. pp. 243–289. ISBN   978-0-07-170078-8 .
• ↑ Daniel R Solin (2012). The Smartest Money Book You'll Ever Read . Penguin Group. pp. 138–141. ISBN   978-0-399-53721-9 .
• ↑ 10.0 10.1 "Beyond the Questionnaire: New Tools for Risk Profiling" . CFA Institute Annual Conference (Montreal). April 27, 2015. Archived from the original on August 15, 2020 . Retrieved May 18, 2023 .
• Portfolio risk management
• Asset allocation

An official website of the United States government

Here’s how you know

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Assessing Your Risk Tolerance

When it comes to investing, risk and reward go hand in hand. The phrase “no pain, no gain” – comes close to summing up the relationship between risk and reward. Don’t let anyone tell you otherwise: all investments involve some degree of risk. If you plan to buy securities – such as stocks , bonds , mutual funds , or ETFs – it’s important that you understand that you could lose some or all of the money you invest.

The reward for taking on risk is the potential for a greater investment return. If you have a financial goal with a long time horizon, you may make more money by carefully investing in higher risk assets, such as stocks or bonds, than if limit yourself to less risky assets. On the other hand, lower risk cash investments may be appropriate for short-term financial goals.

An aggressive investor, or one with a high risk tolerance, is willing to risk losing money to get potentially better results. A conservative investor, or one with a low risk tolerance, favors investments that maintain his or her original investment.

Many investment websites offer free online questionnaires to help you assess your risk tolerance. Some of the websites will even estimate asset allocations based on responses to the questionnaires. While the suggested asset allocations may be a useful starting point, keep in mind that the results may be biased towards financial products or services sold by companies or individuals sponsoring the websites.

Featured Content

Women and Investing

Read our Director’s Take article to learn how women can level the playing field when it comes to investing. 

Investment Professional Background Check Search Tool

Learn about an investment professional’s background, registration status, and more. 

Artificial Intelligence (AI) and Investment Fraud

Read our Investor Alert to learn how bad actors are using AI to lure victims into scams and what you can do to keep your money safe from these frauds.

Investing Quiz – March 2024

This month’s quiz focuses on fraud tactics and scams. Test your fraud literacy today! 

Sign up for Investor Updates