Did you know that most investors don’t know their own returns?¹

It’s not because they don’t read the numbers, but because there’s a widespread misunderstanding about how to calculate your true net returns. Studies have shown, for example, that most people care more about raw returns than anything else. Unfortunately, this isn’t actually the return you should care about, and it can lead you astray.

Only looking at the raw return can skew your perceptions about your portfolio and lead you to make bad choices—such as taking on an inappropriate level of risk or misplacing expectations—which can impede your investing goals.

In fact, there are several important factors to consider when evaluating how your portfolio has performed. This guide will help you reach the right conclusions when judging historical investment returns. By following these steps, you’ll avoid the errors many investors make, and you will correctly infer how well your portfolio has performed, and how it compares to others.

The steps outlined in this detailed primer can be read in order, or you can click down to the topics that are most relevant:

- Total returns, not price returns
- Real returns, not nominal returns
- After tax, not pre-tax returns

- Market risk (beta)
- Manager risk (negative alpha)
- Adjust for cash flows

- Ignore short-term returns
- Annualize returns before comparing them

- Transaction fees
- Expense ratios
- Bid-ask spreads
- Entry/exit fees
- Trading costs

### 1. Use the Right Returns

**Total returns, not price returns**

One of the most common errors investors make is to consider price returns, not total returns. The total return includes not only the change in price of your securities, but also the dividends or coupons they’ve paid out over the time you’ve held the investment.

Figuring out total returns can be a challenge to individual investors, as most news reports and brokerage return figures are based on price returns only. For example, over the past year the S&P 500 has had a price return of 29.7%, but a total return of 32.3%. This implies a dividend return of 2.6%, which is just slightly above historical averages.

Not using the total return when assessing performance means that you could incorrectly penalize investments which have a higher proportion of their total return due to dividends. And remember, reinvested dividends experience the power of compound growth as well.

The graph below depicts the total return and price-only return of the S&P 500 since 1993. The total return growth of a $100,000 portfolio is $514,237 as of January 2, 2014, while the price-only growth is $316,294—a difference of 63%. Put into annual return figures, the price-only return is 8.2% while the total return is 10.2%.

There are six months over this time period when the *price* return was negative, but the dividends pushed the *total* return into positive territory. Thus dividends can make the difference between what appears to be a negative return and a positive return. Just as important, the total return gives you a more accurate sense of how risky or not certain assets were.

**Real returns, not nominal returns**

To get the most accurate returns, you also have to factor in the bite of inflation, especially when comparing returns from different time periods (or different geographies).

A stock that returned 20% in 1920 Weimar Republic, for example, still wasn’t as good an investment as a loaf of bread or a shovel. More relevantly, in 2011 when inflation was an average of 3%, investment returns had to be higher in order to increase wealth in real terms when compared to 2013, with inflation at 1.2%.

In other words, inflation can affect the risk/return profile of any asset, including cash. Money stashed under the bed is virtually certain to lose value over time. Savings accounts seem to provide a certain guarantee that equities and bonds don’t. In reality, while savings accounts may guarantee your nominal investment, they are also guaranteed to steadily erode your real wealth over time.

By including inflation, you can pinpoint real returns, which is a more accurate gauge of growth than nominal returns.

**After tax, not pre-tax returns**

Finally, most investors care about the growth in wealth which they (as opposed to the government) actually benefit from. But estimating your net-of-tax return can be tricky, as the short-term capital gain you realize in February won’t be taxed till the following April—but it will be taxed. As such, you should consider your portfolio’s returns after you’ve deducted the amount of tax you’ll have to pay on it. (Note that this only applies for taxable portfolios.)

There are generally three sources of tax-drag on your portfolio:

1. **Ordinary income tax vs. capital gains tax:** Although qualified dividends (generally from large, U.S. companies) are taxed at a reduced rate, other income generated by your portfolio, be it bond interest or non-qualified stock dividends, will be taxed at your highest marginal income tax rate. Given that your income tax rate will generally be the same as, or higher than, your capital gains tax rate, it’s important to bear in mind that the same dollar return is likely to give you a higher after-tax return if it comes from capital gains rather than other investment income.

2. **Short-term capital gains vs. long-term capital gains:** Capital gains on investments which you’ve sold after holding for a year or less (short-term) are taxed at a higher rate than those you’ve held for more than a year (long-term). Strategies which realize short-term capital gains are therefore penalized up to an additional 20% on such gains, versus strategies that favor holding for more than 12 months. This is a major reason why frequent trading can lose any lustre once returns are calculated on an after-tax basis. Those costs may be invisible until next April, but they are there.

3. **Taxes now versus taxes later:** In general, the longer you defer taxes, the longer that money you’ll eventually part with can work for you in the meantime. Even long-term capital gains are taxed when you realize them, and so avoiding realizing capital gains has its own benefit—greater tax-deferred growth. Strategies which defer taxation can therefore be better, all else being equal, than those which realize capital gains. If you realize your gains at a point when you have a relatively low marginal tax rate (if you take a year off, or in retirement), you could owe as little as zero capital gains tax, meaning you pay zero tax on that portion of your returns.

### 2. Adjust for Risk

When comparing two different investments, you must consider how much risk you took on to achieve the returns of each.

Consider two investors, Jake and Debbie. Jake achieved a return of 7% this year, and Debbie achieved 6%. You might say that Jake performed better than Debbie, but Jake had a very risky portfolio, in which he had a chance of losing more than 50% of his investment, and had to endure a bumpy ride throughout the year. Meanwhile, Debbie had a zero-risk investment with no drawdowns.

In other words, Jake bore a lot of risk and stress to earn a 1% higher return. So in fact Debbie’s portfolio performed better—i.e. had higher *risk-adjusted returns*—than Jake’s did.

There are generally two sources of risk you should consider—market risk, and idiosyncratic manager risk.

**Market risk (beta)**

Market risk is uncertainty and volatility due to the financial markets as a whole.

Most stocks are all fairly correlated with each other—whether or not expectations for the world economy are positive or negative. This single source of risk—or market beta—drives the vast majority of stock returns, and most stocks are subject to it.

Thus, the amount of stocks in your portfolio will determine how sensitive it is to beta. So the first thing you should do is consider how much of your portfolio’s return was due to beta. It’s not correct to compare a portfolio with a beta of 90% to a portfolio with a 70% beta unless you adjust for the fact that one has a higher risk exposure than the other.

To compare returns with different levels of risk, econometricians use a metric called a Sharpe ratio, defined as the excess return² of the portfolio divided by the standard deviation of the return (the volatility). Thus, if two portfolios receive the same return in a given time period, the one with the lower volatility will be awarded the higher Sharpe ratio—it has achieved the higher return with greater certainty.

**Manager risk (negative alpha)**

While investing in a well-diversified market portfolio means that you have a controlled exposure to market risk, many investors use managers who actively deviate from holding a market portfolio. This often means you hold a subset of stocks or bonds which reflect the manager’s views. As a result, they are taking on *idiosyncratic manager risk*, i.e. the greater variance in outcomes associated with using an active manager. This risk is independent of the beta in your portfolio.

As we discussed above, you should only take on more risk when you expect higher returns in exchange. Unfortunately, as we found in our white paper with Rick Ferri, *A Case for Index Fund Portfolios*, the expected return of bearing manager risk is *negative*. Active mutual funds are riskier, and have lower expected risk-adjusted returns than a passive portfolio.

Another key insight from this white paper is that the time frame matters. Over a shorter period, you may conclude that a given manager has skill. But over a one-year period, odds are just over 50% that an active manager will underperform the index. Over five years, that rises to 77%, and over 15 years, 83%. So one year of good performance is a weak indicator that the manager is a long-term winner.

When assessing your portfolio’s performance, and comparing to an active manager, it’s important to consider this manager risk factor. You took an extra chance by using a manager—you should be consistently rewarded with above-average returns for that risk. If your manager doesn’t outperform consistently, it’s probably not worth the risk.

**Adjust for cash flows**

Finally, we should mention that you, the investor, are also a manager. You could be making deposits and withdrawals over the course of the year, and these cash flows mean that you may experience different overall returns than a buy-and-hold portfolio. Significant amounts of cash flows may result in a positive or negative behavior gap, i.e. the difference between the returns of the market (assuming a buy-and-hold strategy), and the actual returns you the investor received, based on your behavior.

Research has shown that the riskier the fund, the greater the behavior gap. While we can’t completely prevent bad behavior, Betterment’s customers have the lowest measured behavior gap we know of.

### 3. Adjust for Time

The period of time you’ve been invested is perhaps one of the trickier but more important adjustments to be made. The two things to keep in mind are to not trust short-term returns, and always annualize returns to compare like-for-like.

**Ignore short-term returns**

A very liberal definition of “short-term” means anything less than three years of monthly return data (36 observations) just to look at an investment on its own. A conservative definition is about 12 years. And If you want to compare two investments, you may need twice that amount of time.

While we can’t completely prevent bad behavior, Betterment’s customers have the lowest measured behavior gap that we know of.Why shouldn’t you trust short-term returns? The shorter the period of time, the higher the chance you have of coming to the wrong conclusion. There is too much randomness in short-term returns to make them reliable. It’s very easy to have a diversified portfolio be beaten by any of its constituents in the short-term—only to beat all its constituents in the long-term. Looking at short-term returns is like performing a statistical test with very little data—you just can’t have any confidence in the results.

**Annualize returns before comparing them**

Second, you should always annualize returns. For example, imagine you had a two-year return of 25%, and were comparing it to a three-year return of 30%. Which one is better? Clearly, the 30% is larger, but it’s also had more time to grow. What would the 25% return be equivalent to over three years?

If you annualize the returns using the equation below, you’ll find that in this case, the two-year return is equivalent to an annual return of nearly 12%. The three-year return is equivalent to an annual return of 9%. So adjusting for the amount of time you’ve been invested, the two-year actually has better performance.

**Annual return** = ((1 + return / 100) ^ (1 / years invested) – 1) x 100

**Two-year example:** 11.8% = ((1 + 0.25) ^ (1 / 2) – 1) x 100

**Three-year example:** 9.1% = ((1 + 0.30) ^ (1 / 3) – 1) x 100

### 4. Adjust for Costs

Every investment has its own costs that an investor must pay in order to get access to its returns. Therefore, any comparison of returns that fails to account for all costs associated with the relevant investment products is misleading.

Let’s take a look at some of the costs most commonly encountered by individual investors:

**Transaction fees:** This is money you pay to trade a security, and you’ll usually pay about $5-$20 for a discount broker, and often $150 or higher for a full-service broker. Given that you’ll sell at some point as well, consider these costs doubled. Moreover, as you accumulate wealth, you’re likely to want to add to your holdings, and $10 every time you want to put a little away can really add up.

There are some brokers that offer some free ETF trades, but these carry limitations—typically only a limited number of trades are allowed, and only a limited range of ETFs are available.

**Expense ratios:** This is the total annual amount of a fund’s assets that is used to cover its operating costs—including administrative, management and advertising costs. Both mutual funds and ETFs have expense ratios, but returns are normally stated net of this cost. One thing to remember, however, is that indices, such as the S&P 500, are not actual investable assets. Thanks to great innovations such as Vanguard’s VOO, you can get very close to true S&P 500 returns—but you’ll still have to pay an expense ratio for an index fund that tracks the S&P.

**Bid-ask spreads:** This charge is also a trading cost of sorts; you can think of it as money lost to friction each time you trade. The bid-ask spread is simply the difference between the ‘buy’ and the ‘sell’ price available at market. For the most liquid securities this is normally only a few hundredths of a percent, but can be much higher in extremely volatile or illiquid markets. For micro-cap stocks, for example, the bid-ask spread can be huge, and stocks may have to increase by several percent—or even much more—before they can be sold at a profit.

**Entry/Exit fees:** Also known as sales fees, load fees are particularly nasty charges that are commonly 5% of assets or more. You pay a front-end load when you buy the fund, or a back-end load when you sell it. Here’s the real kicker: Current research shows that mutual funds with a load fee actually underperform those with no load fee, even taking into account the loads themselves. Thus, load fees should be avoided at all costs, and if your advisor or broker recommends one to you, you should think about whether their incentives are conflicting with yours.

Other fees to consider are advisory fees, subscription fees, or brokerage maintenance fees that aren’t tied to specific assets and so can’t reasonably be included in returns when those are reported. But if you’re paying those costs, they are nipping into your returns. Remember: any percentage you save in these costs will 100% go towards increasing your returns.

### Your returns, the full picture

When you add up all the points above and calculate how, in concert, they all may have an impact on your returns, you can see why looking at just your portfolio’s raw returns is not an effective way to understand performance. And while it may seem that these numerous factors would mostly eat into your final returns, the more useful perspective here is to remember that your focus—and ours—is to optimize investing returns and inoculate your portfolio’s growth against the potential hits from taxes, idiosyncratic risk, costs, and so on.

By being aware of each of these and how they operate you can better protect your returns and feel more confident about reaching your investing goals.

*Patrick Burns contributed to this post.*

¹*Glaser, Markus and Weber, Martin, *Why Inexperienced Investors Do Not Learn: They Do Not Know Their Past Portfolio Performance* (November 15, 2007). Finance Research Letters, Vol. 4, No. 4, 2007.*

²*An excess return is the return above and beyond that achievable from investing in a risk-free asset. For instance, if a portfolio has an 8% return over a period of time when the risk-free rate is 2%, the portfolio’s excess return is 6%.*

Hi Dan, excellent display! Much thanks.

So is this how the returns are displayed on when using betterment? or is this something we still need to figure out by hand?

Hi Joel, Thanks for your comment.

The returns as displayed in your Betterment account are net of all costs, cash flows and reflect total returns, rather than price returns. Our forward projections that supply the data to our advice engine take into account expected risk, costs, cash flows and time. However, there are some elements mentioned in the post that are just not possible to display.

Because everyone’s tax situation is different, the returns we show cannot reflect their taxes without a complete picture. We make after-tax returns as easy as possible to calculate by integrating with great products such as TurboTax and by providing generic tax information, but even then the full implications of your tax situation won’t be realized until the end of each tax year. That’s not to say you can’t make estimates as you go along.

As for inflation, time and historic risk – these are important things to consider when interpreting returns in your Betterment account or comparing two different returns – but they’re not something we can really display in a way that makes sense for all customers in all situations. You can, however, look at returns over different time periods on the analysis tab in your Betterment account.

Hope that clarifies some of your questions.

So, I get adjusting for actual returns, time, and costs – but how do you adjust for risk? Like, in that scenario, how much more return would Jake have to have to be considered equal to Debbie? Once you’ve measured risk, how do you incorporate it into your returns to know your “risk-adjusted” returns?

There isn’t enough information in the example to calculate that analytically, and the example itself is a little imprecise in saying, “he had a chance of losing more than 50% of his investment.” ANY portfolio has a chance of losing 50%, or even 100%, of the investment; what that chance is depends on the portfolio and the time frame and is defined in terms of the confidence interval.

In investment discussions the confidence interval is typically 2 standard deviations, or 95% of expected outcomes, over the next year, but market returns aren’t always normally distributed. The direct answer to your question is, for the portfolio A with excess return R and volatility X to be risk-equivalent (under the Sharpe Ratio definition) to portfolio B with volatility 2X, the expected excess return of B must be 2R. Excess return is return over a risk-free portfolio, as explained in footnote 2. The usual proxy for risk-free rates is US Treasury yields, which until recently were near zero. As also noted in the article, Treasuries are subject to inflation risk, and also reinvestment risk (the risk that at maturity, yields, and hence returns, are lower), but the usual definition of risk-free encompasses loss of principal, as from default.

There are other ways to measure risk-adjusted returns, e.g. the Sortino Ratio, which is similar to Sharpe except the denominator is only downside volatility, and the Information Ratio, which measures excess return (alpha) over the standard deviation of the excess returns (tracking error). None of them are necessarily that intuitive, but as Dan wrote above, the vast majority of actively managed portfolios are less optimal from a risk-return standpoint. For most investors, the optimal risk-return opportunity is some weighting of fixed income and equity index portfolios along the efficient frontier. See the blog posting about the new portfolio for an illustration of that.

Great post and great service! One comment on the annual return formula: (1 + 0.25) ^ (1 / 2) is approximately 1.118. You need to subtract 1 then multiply by 100 to get to 11.8.

Thanks for the comment, Joshua!

You’re certainly right – we’ve updated the post to make that more clear.

Great post, I definitely needed this information. Thanks for sharing!