Investment Risk and Return: The Math Behind Your Money

Investment risk and return belong in the same conversation. A stock might deliver an impressive long-run average, yet a sharp loss just before you need the money could derail your goal. Meanwhile, an apparently steady investment might quietly lose purchasing power to inflation. To compare your choices well, you need to understand what a return measures, how much returns vary, and what happens when gains and losses compound.

This guide turns the essential math into practical questions. You will learn how to calculate returns and volatility, why the bell curve can mislead investors, when to use arithmetic or geometric averages, and how to evaluate a portfolio against your own time horizon. The examples are educational and use hypothetical numbers unless a historical source appears beside them.

What Do Investment Risk and Return Mean?

An investment’s return describes what you gain or lose relative to the amount you invested. Its risk includes uncertainty about that outcome and the possibility of losing money. According to the U.S. Securities and Exchange Commission’s investor education site, investors generally seek higher expected returns when they accept greater risk. However, greater risk never guarantees a greater realized return.

For example, consider two hypothetical investments. One pays a fixed 4% over a year if its issuer meets its obligations. Another might gain 25%, remain flat, or lose 30%. The second offers more upside, but your actual result depends on what happens during the year. Moreover, the apparently predictable first investment still faces inflation, issuer, and possibly reinvestment risks.

Total return includes income and price changes

For a single holding period, start with:

Holding-period return = (ending value − beginning value + cash income) / beginning value

Suppose you buy a share for $100, receive a $3 dividend, and sell the share for $108. Your total return is ($108 − $100 + $3) / $100 = 11%. If you ignore the dividend, you understate your result. Conversely, if you compare a dividend-paying fund with a price-only index, you may compare different things.

For a multi-year fund comparison, check whether both return series include reinvested dividends, distributions, expenses, and the same dates. The SEC’s performance-claims bulletin specifically urges readers to examine calculation methods, fees, benchmarks, and cherry-picked periods.

Nominal return differs from real return

A nominal return measures the change in dollars. A real return adjusts for changes in purchasing power. If an investment rises 6% while consumer prices rise 3%, the exact inflation-adjusted return is:

Real return = (1 + nominal return) / (1 + inflation rate) − 1

Thus, 1.06 / 1.03 − 1 = 2.91%, approximately. Subtracting 3% from 6% gives a convenient 3% estimate, but the ratio gives the exact relationship. The Bureau of Labor Statistics explains how the Consumer Price Index helps measure the changing purchasing power of a dollar. Your personal inflation rate may differ from the broad index because your spending differs from the average basket.

Why Do Investors Expect a Reward for Risk?

People usually prefer a more certain outcome when two choices offer the same expected payoff. Therefore, risky assets often need to offer the possibility of a higher return to attract buyers. This expected compensation is a risk premium, although investors only discover the actual premium after the fact.

Historically, broad U.S. stocks have generally earned more than Treasury bills over long periods. For a transparent annual series that includes stock dividends, see Aswath Damodaran’s historical stock, Treasury bond, and Treasury bill data. That historical pattern does not mean stocks win every year or during every investor’s holding period. FINRA’s discussion of risk also cautions that a long holding period does not eliminate the possibility of a damaging loss near the date you need your money.

Different risks affect different investments

• Market risk: A recession, crisis, or change in sentiment can lower prices across many assets.
• Company risk: A single business can disappoint, lose market share, or fail.
• Interest-rate risk: Rising rates can reduce the market value of existing fixed-rate bonds.
• Credit risk: A borrower may miss promised payments.
• Inflation risk: Your nominal gains may buy fewer goods and services than expected.
• Liquidity risk: You may struggle to sell quickly at a fair price.
• Currency risk: Exchange-rate movements can change the return on foreign assets in your home currency.
• Sequence risk: Poor returns just before or during withdrawals can cause more damage than the same losses earlier in an accumulation period.

Standard deviation, which we will calculate below, captures the variability of returns. However, it cannot summarize all these risks. A bond, for instance, might show modest historical price variation while still exposing you to inflation or default. Likewise, an investment with a smooth reported price can carry risks that become visible only when you try to sell.

What Does Historical Volatility Tell You?

Volatility describes how widely returns fluctuate. Analysts often measure it with standard deviation. A higher standard deviation means returns have tended to move farther from their average; it does not automatically tell you whether the next move will be up or down. FINRA describes volatility in mathematical terms as the standard deviation of returns.

Imagine that Investment A returns 4%, 6%, 5%, 5%, and 5% over five years. Investment B returns −20%, 30%, 5%, 25%, and −20%. Both have a 5% arithmetic average. Nevertheless, their paths feel very different, and Investment B’s losses make its compound outcome worse.

How to calculate the average, variance, and standard deviation

For annual returns r₁ through rₙ, the arithmetic average is:

Average return = (r₁ + r₂ + … + rₙ) / n

Next, subtract the average from each return, square each difference, add those squares, and divide. If your observations are a sample from a wider return process, the usual sample variance divides by n − 1; the population variance divides by n. Finally, take the square root to get standard deviation in the original units of the returns.

Sample standard deviation = √[Σ(rᵢ − average return)² / (n − 1)]

Consider hypothetical annual returns of −10%, 0%, and 20%. Their arithmetic average is (−10% + 0% + 20%) / 3 = 3.33%. The sample standard deviation is approximately 15.28 percentage points. A spreadsheet’s STDEV.S function would calculate the sample statistic; STDEV.P uses the population denominator. As a result, two analysts can report slightly different volatility for the same short series if they choose different conventions.

Also match the time interval. Daily, monthly, and annual standard deviations are different quantities. Analysts commonly estimate annualized volatility from monthly observations by multiplying monthly standard deviation by √12, but that shortcut assumes the relevant dependence structure supports the conversion. It does not make next year’s returns predictable.

Why a single average hides the investment experience

An arithmetic average compresses many outcomes into one number. Therefore, always ask for the period, the range of outcomes, and the sequence of returns. Five calm years and five turbulent years can share the same arithmetic average even though their account balances end at different values.

In addition, investment risk depends on when you will need the funds. A 25% decline can inconvenience someone investing for decades and devastate someone making a down payment next month. The SEC’s asset-allocation guide ties a suitable asset mix to the investor’s time horizon and tolerance for loss.

Does the Bell Curve Describe Stock Market Returns?

A normal distribution is a symmetrical bell-shaped model described by a mean and a standard deviation. Under a true normal model, about 68% of observations fall within one standard deviation of the mean, about 95% within two, and about 99.7% within three. These are properties of the model, not reliable promises about future market returns.

Suppose, purely for illustration, that an asset’s annual mean is 8% and its standard deviation is 20 percentage points. A one-standard-deviation interval runs from −12% to 28%. Under an exact normal assumption, roughly two-thirds of years would land within that interval. However, that sentence depends entirely on the assumed distribution and stable estimates. It does not say an actual investor has a two-thirds chance of earning those results next year.

Why the normal model can understate crashes

Financial returns can show heavier tails than a normal curve, meaning extreme outcomes may occur more often than that simple model suggests. For example, research published through the Bank for International Settlements discusses how risk calculations based on normal returns can miss the fat-tailed behavior of actual returns. Moreover, volatility changes through time, and falling markets can coincide with rising correlations among holdings.

There is another mathematical warning: a normal distribution for simple annual returns allows values below −100%, although a fully paid, unleveraged investment cannot lose more than its initial value. Consequently, a normal curve may serve as a classroom approximation, but you should never treat its outer tails as a realistic complete map of investment outcomes.

Better questions than “What is the probability of a normal year?”

Ask how the investment behaved during major downturns, what its worst historical drawdown was, and whether you could meet your goal after a large loss. In addition, inspect the assumptions behind any forecast or probability figure. A model can clarify a decision, yet its neat percentages can obscure unfamiliar risks.

Why Do Big Losses Require Even Bigger Gains?

A percentage gain applies to the money remaining after a loss. If $100 falls by 50%, you have $50. A subsequent 50% gain adds $25, leaving only $75. To get back to $100, the remaining $50 must double.

The required recovery gain after a loss of L, expressed as a decimal, is:

Required gain = L / (1 − L)

For example, a 75% loss reduces $100 to $25. Therefore, the account needs a $75 gain on its new $25 base, or 300%, to break even. This arithmetic helps explain why controlling severe losses matters, even when an asset has attractive upside.

Dollar losses and percentage losses answer different questions

Suppose an index falls 500 points from 2,000. The decline is 25%. If it falls the same 500 points from 20,000, the decline is only 2.5%. Thus, a headline about the “largest point drop” says little about its size relative to the starting level. Use percentage change = (new level / old level − 1) × 100 to compare different eras.

Arithmetic vs. Geometric Average Return: Which One Should You Use?

The arithmetic average describes the simple average of periodic returns. The geometric average describes the constant per-period rate that would produce the same ending wealth as the observed sequence. Investors also call the latter a compound annual growth rate, or CAGR, when the periods are years.

These measures answer different questions. If you ask, “What was the average of the individual yearly returns?” calculate the arithmetic mean. If you ask, “At what annual rate did my original investment actually compound over this period?” calculate the geometric mean.

A two-year example: 50% down, then 100% up

Start with $100. After a 50% loss, you have $50. A subsequent 100% gain returns the balance to $100.

• Arithmetic average: (−50% + 100%) / 2 = 25% per year.
• Geometric average: √[(1 − 0.50)(1 + 1.00)] − 1 = 0% per year.

The account finished where it started. Therefore, 0% describes its actual annual compound growth, even though the arithmetic average is 25%. Neither calculation is wrong; each addresses a different question.

How to calculate the geometric average

For n yearly returns, convert each percentage to a decimal and compute:

Geometric average = [(1 + r₁)(1 + r₂) … (1 + rₙ)]^(1/n) − 1

Suppose annual returns are 10%, 12%, 3%, and −9%. Multiply 1.10 × 1.12 × 1.03 × 0.91 = 1.1547536. Thus, a $100 investment grows to approximately $115.48, assuming no external contributions or withdrawals. The geometric average is approximately 3.66%, while the arithmetic average is 4%. Small differences compound over long periods.

If an investment reaches zero, you lose the entire starting balance and cannot recover through later percentage gains on that same zero balance. In that case, its geometric rate over the period is −100%, provided you include that total loss in the series.

Why the geometric average usually trails the arithmetic average

For the same positive ending-wealth factors, the geometric mean cannot exceed the arithmetic mean. In practice, volatile returns create a wider gap because losses reduce the capital available for subsequent gains. If each annual return happens to be identical, both means match.

Analysts sometimes approximate the gap for moderate variability with geometric mean ≈ arithmetic mean − variance / 2, using returns as decimals. However, this is an approximation under specific assumptions, and it becomes unreliable for extreme returns or unusual distributions. Calculate the geometric mean directly when you have the actual sequence.

Which average belongs in a forecast?

For a description of past compounded wealth, use the historical geometric average. For the expected return of a single future period under a specified probability model, the arithmetic expectation usually answers the mathematical question. A multi-year forecast is harder: uncertainty, changing conditions, and estimation error can make either historical number a poor prediction.

Marshall Blume’s 1974 study on estimating long-run expected rates of return examined biases that arise when analysts use sample arithmetic or geometric means for longer-horizon forecasts. Some textbooks present a weighted adjustment between the two. Treat any such adjustment as a model-dependent estimate, rather than a rule that turns old returns into a dependable forecast.

What If You Add or Withdraw Money Along the Way?

When cash enters or leaves your account, the account’s beginning and ending balances alone do not tell the full performance story. A portfolio’s time-weighted return measures the performance of the investments across subperiods while reducing the effect of external cash-flow timing. A money-weighted return, often calculated as an internal rate of return, reflects the timing and size of your deposits and withdrawals.

For example, imagine that an account rises from $100 to $150, you deposit another $100, and then the account loses 20%. The final balance is $200. Because your total contributions also equal $200, you might think the investments earned nothing. Yet the investment path itself produced a 20% cumulative time-weighted gain: 1.50 × 0.80 − 1 = 20%. Your large deposit arrived immediately before the loss, so your personal dollar result differs from the investment’s chained return.

Therefore, use time-weighted figures when judging a manager’s investment performance, and examine money-weighted results when assessing what you earned on the dollars you actually invested. Also verify whether performance figures account for taxes and fees. The SEC explains that fees reduce the capital left to compound.

What Can Historical Returns Tell Us About Stocks, Bonds, and Cash?

Long histories provide context about what happened under many economic conditions. Damodaran’s U.S. historical-return dataset, for instance, separates stock total returns, three-month Treasury bills, and returns on ten-year Treasury bonds. It also distinguishes a bond’s total return from its quoted yield: when prices move, the return from holding a bond can differ from its yield at the start of the period.

However, a historical table is not a menu of guaranteed future returns. Its answer depends on the chosen starting year, ending year, asset definition, dividend treatment, and whether it reports nominal or inflation-adjusted results. A small-stock portfolio from one old textbook also may not match a modern small-cap index. Accordingly, cite a date range and the exact series whenever you publish a historical average.

Stocks

Broad stocks give investors a claim on company earnings and can participate in long-run business growth. Nevertheless, share prices can decline sharply and stay below earlier peaks for years. Company-specific risk becomes even more important when you hold only a handful of stocks.

Bonds

Bonds may provide scheduled interest payments, but their market prices can change. A bond or bond fund with a longer duration generally responds more strongly to interest-rate moves. In addition, issuers can differ greatly in credit quality. Do not assume every bond acts like a short-term Treasury bill.

Cash and Treasury bills

Cash-like assets can support near-term goals because their nominal balances generally fluctuate less than stock prices, depending on the product. Still, a lower nominal return can translate into a negative real return during inflation. Moreover, renewal rates can change, so today’s yield does not lock in future yields indefinitely.

How Does Diversification Change Investment Risk and Return?

Diversification spreads money across holdings that do not always react identically. If one company fails, other holdings may soften the portfolio’s overall loss. Combining asset types can also change the balance between expected return and volatility. The SEC’s diversification guidance says diversification can reduce risk, while warning that it cannot guarantee protection when markets fall.

Diversify within and across asset categories

Owning 20 companies from the same industry can leave you exposed to one common shock. Conversely, a broad stock fund may spread company risk across many businesses, while a suitable bond or cash allocation can address a different set of needs. Your mix should follow the purpose of the money, not just the number of tickers in your account.

Correlation matters too. Two assets with the same individual standard deviation can create a steadier combination if their returns do not move in lockstep. However, relationships can change during crises, so a historically low correlation does not promise future protection.

Match your portfolio to the date of your goal

Money needed soon generally has less time to recover from a large market decline. Meanwhile, an investor saving for a distant goal may choose to accept more short-term fluctuation, depending on their capacity and willingness to bear losses. The SEC’s guide to asset allocation and rebalancing explains why the right mix may change as a goal approaches.

After you choose a mix, rebalancing can restore the weights you intended when market moves push them out of line. First, check transaction costs and potential taxes. Then decide whether a calendar schedule or a preset drift threshold fits your plan.

How to Compare Two Investments Without Being Misled

Before choosing a fund, strategy, or asset allocation, work through these questions:

1. What does the return measure? Check total return, dividends, reinvestment, the time period, and the currency.
2. Which average does the headline use? Ask for geometric growth when evaluating what a lump sum actually became.
3. How large were the setbacks? Examine annual losses, maximum drawdown, and the time it took to recover.
4. Are the risks comparable? Consider concentration, credit quality, duration, liquidity, and foreign-currency exposure.
5. Does the comparison include costs? Fees, trading, and taxes can change the result you keep.
6. What did inflation do? Compare purchasing power when the goal involves future spending.
7. Is the benchmark appropriate? A U.S. large-company stock index may be a poor comparison for a global bond fund.
8. Did someone select a flattering window? Review more than one period and include weaker markets.
9. When will you need the money? An attractive long-run average cannot pay a bill after a short-run crash.

The SEC’s bulletin on performance claims supports these checks and warns that historical performance cannot predict future results.

A quick spreadsheet workflow

Place annual total returns in cells B2:B11, expressed consistently as percentages. Then use:

• =AVERAGE(B2:B11) for the arithmetic average.
• =STDEV.S(B2:B11) for the sample standard deviation.
• =GEOMEAN(1+B2,1+B3,1+B4,1+B5,1+B6,1+B7,1+B8,1+B9,1+B10,1+B11)-1 for geometric growth where the spreadsheet supports that syntax and every growth factor is positive.

Alternatively, create a helper column with =1+B2 through =1+B11, and use =GEOMEAN(C2:C11)-1. That approach makes your assumptions easy to audit. To reconstruct ending wealth from a $100 starting amount, calculate =100*PRODUCT(C2:C11). If your series includes a −100% return, the final wealth is zero; handle that case separately because a zero factor prevents further growth of the original investment.

Be careful with units: a cell containing 10% already stores 0.10, while a cell containing the number 10 does not. Also avoid taking the average of a fund’s annual returns and presenting it as the annual compound growth rate.

Frequently Asked Questions About Investment Risk and Return

Is a higher average return always better?

No. First identify which average the source uses. Next, compare losses, fees, taxes, liquidity, inflation, and the date you will need the funds. A return that looks attractive over decades may be unsuitable for a goal next year.

Is standard deviation the same as the chance of losing money?

No. Standard deviation measures dispersion around an average. To estimate a loss probability, you need a credible model of the whole return distribution, and actual markets may depart from that model. Additionally, standard deviation treats unusually high gains as deviations too, although investors may welcome them.

Can I assume that 95% of returns will fall within two standard deviations?

Only if a suitable normal model applies and its mean and standard deviation remain relevant. Actual financial returns can have heavy tails and changing volatility. Therefore, do not turn the bell-curve rule into a promise about any specific stock or future year.

Why is my CAGR below the advertised average annual return?

The advertisement may use an arithmetic average of yearly percentages, while CAGR reflects the compounding of your starting investment. Fluctuating returns usually make the geometric figure lower. Fees, differing dates, and cash flows may widen the difference further.

Can a diversified portfolio lose money?

Yes. Diversification can reduce exposure to a single holding and reshape portfolio risk. However, broad markets can fall together, and no mix eliminates every risk. You still need an allocation suited to your goal and your capacity for losses.

Are past returns useful if they cannot predict the future?

Yes, if you use them carefully. Historical returns show how an asset or strategy behaved in earlier conditions, reveal volatility and drawdowns, and help you test assumptions. Nevertheless, a past average should serve as context for a decision, not as a guaranteed growth rate.

Conclusion: Use the Right Number for the Right Decision

Investment risk and return become easier to evaluate when you separate four questions: What happened to the investment, how widely did results vary, what happened to your compounded wealth, and when will you need the money? Arithmetic averages describe individual-period returns; geometric averages describe historical compound growth. Standard deviation summarizes variation, while drawdowns, inflation, fees, and your time horizon fill in crucial gaps.

Before acting on a headline return, look at the full path and the assumptions behind the number. Then choose a portfolio you can hold through plausible setbacks while still meeting the purpose of your money.

Sources and Further Reading

• U.S. Securities and Exchange Commission, What Is Risk?
• U.S. Securities and Exchange Commission, Investor Bulletin: Performance Claims
• U.S. Securities and Exchange Commission, Beginners’ Guide to Asset Allocation, Diversification, and Rebalancing
• U.S. Securities and Exchange Commission, How Fees and Expenses Affect Your Investment Portfolio
• FINRA, Risk
• U.S. Bureau of Labor Statistics, Consumer Price Index Frequently Asked Questions
• Aswath Damodaran, New York University Stern School of Business, Annual Returns on Stock, T. Bonds and T. Bills
• Bank for International Settlements, Comparative Analyses of Expected Shortfall and Value-at-Risk
• Marshall E. Blume, Unbiased Estimators of Long-Run Expected Rates of Return

Share on Facebook

Post on X

Follow us

Save