Amongst all the edicts investors should heed, one stands out above all others: It’s time in the market that builds returns, not market timing.
We’ve illustrated this rule with one of the longest-running index data sets available: the daily returns of the S&P 500 from Jan. 2, 1928, through 2014. You can explore the difference that a few years makes by changing the slider or pressing the play button. The interactive presents the distribution of returns an investor would have received for an investment of $100 for a given holding period in the S&P 500.
Returns by Various Investment Periods in the S&P 500
Historical probability of a loss and median return by period length
How to Read the Interactive
x-axis: This represents the final value of a $100 investment over a given period in the S&P 500. To change the investment period, use the slider in upper left corner.
y-axis: This percentage shows the historical proportion of outcomes. For example, 99.3% of investment periods lasting one day resulted in a final investment value between $95 and $105.
Example: From 1928 to 2014 there were 21,502 possible holding periods that lasted 12 months (e.g., Jan. 2, 1928, to Jan. 2, 1929; Jan. 3, 1928, to Jan. 3, 1929, etc.). Of those 21,502 holding periods, 74% resulted in positive returns with a median return of $13 on a $100 investment. On the other end of the slider, there were 18,730 possible holding periods that lasted 12 years. The returns for these holding periods were more widely dispersed and overwhelmingly positive, with a median return of $240 on a $100 investment.
Risk in the Short Term
The graph below focuses on the relationship between probability of loss and holding period. After one year, the proportion of losses drops substantially. This is one key part of the time-in-the-market advice. If investing in the S&P 500, the risk of loss is much higher for the short-term investor than the long-term investor.
Holding Period and Probability of Loss in the S&P 500
Return in the Long Term
Conversely, let’s look at the median total cumulative return for every given holding period. It’s a stable and steady line, rising over time.
Holding Period and Cumulative Return in the S&P 500
In combination, we can see clearly the imperative to keep the focus on time. The broader market is, and will almost always be, riskier for the short-term investor than the long-term investor. We previously wrote about the benefit of time in the market and maintaining a long term focus during downturns. We hope this analysis bolsters those points.
Lastly, given the importance of time in the market, Betterment tailors your stock allocation to match your personal investment timeframe. Learn more about how our goals work with your timeframe.
This blog post was produced with Joe Jansen.
Thanks Dan for this article and it certainly matches with my investment experience as well till now. (till now being the qualifier :-))
The one comment/question that I had regarding this analysis/study is, how would the investment returns or probability of loss vs gains changes when dollar cost averaging on weekly or monthly basis is taken into account? Does this skew the results in a particular direction?
Hi Ashok,
It’s a good question.
The core driver doesn’t change – the chance of a loss goes down over time since markets go up over time. Having many different deposits at different times just means you have a much more complicated weighted average to calculate. We’ll actually have another post on this issue – time weighted versus money weighted return calculations – very soon.
But the fundamental rule applies across all the individual deposits – they longer they are invested for, the greater the chance of a gain.
This is part of why it’s not worth holding money back out of the market – you’re just spending time out of the market, which is the key driver of your profit or loss.
Best,
Dan
Dan, thanks for this article. I am now better equipped, information wise, to convince friends and family to invest for the long haul. Keep these types of articles coming, please!
Thanks Dan, interesting study. Not clear if inflation was included in the data, but if it was not, how would inflation affect the number of negative outcomes?
Hi Manhattan,
See my answer below.
Best,
Dan
Why not use betterment’s mix of stock ETFs instead of the S&P 500?
Hi Kyle,
Good question.
The first reason is that I wanted to make the point in the most general way possible. Betterments portfolio actually has better historical risk-adjusted returns than the S&P500, but the point about investment period leading to a higher chance of gains isn’t specific to our investments.
The second reason is data. Historical proxy data on our portfolio goes back to 2000 at the earliest. For this analysis, which ends up looking at many different 10 year periods, I wouldn’t have nearly enough data if I used our portfolio. So the longevity of data on the S&P 500 wins in this case.
Best,
Dan
Thanks for the article and the interactive chart. I am wondering, are those numbers by any chance inflation-adjusted? Because, at the end of the day, a nominal dollar amount doesn’t say much about how much you can do with that money.
Are the final values in the interactive inflation-adjusted?
Hello Manhattan, Andreas, and Kristin,
Great point. These numbers were not adjusted for inflation. I might add inflation adjustment as an option in the future.
While inflation would definitely reduce returns across the board it doesn’t effect the overall story – the longer you invest for, the higher the chance of seeing larger gains.
Best,
Dan
My question is that I get confused about profit and loss displayed as average returns, cumulative returns or actual returns. If you have $100 and you lose 50% you have $50. If you then have a 50% increase many people think you are back to $100 but actually you have $75 so it can often be misleading to speak about percentage gain and loss without talking about actual dollar value. Does your presentation address this?
Hi Ray,
Yes, it does. The x-axis is always the final value of the $100 dollar investment.
Best,
Dan
Dan, i’m a newbie investor. I put in some money in betterment just to let it grow. My question is, does the “earned” money get automatically reinvested or not? Does compounding interest work here?
Yes, it does. Every dollar is reinvested automatically.
Dan, great piece. In addition to inflation, which I understand you are going to add, I suggest you also add data to show not only positive returns (i.e. returns higher than zero, post inflation) but also returns higher than, say, 5% per annum. If I invest in equities it’s because in the long run I hope to get at least that…otherwise I might as well invest in fixed income stuff! And…when are you guys launching in the UK?
Hi Luigi,
Good point. Rather than an arbitrary return however, we’ll compare the return on stocks to the risk-free rate in the same period, to see what you could have gotten ‘risk free’.
Dan
This is not directly related to the point of this article, but I assume that this article was prompted by recent drops in indexes (especially in EU) and Betterment accounts, and my question is related to that.
I get the reference to a study that indexes beat actively managed funds, but how can we address the fact that sometimes indexes enter range-bound periods and do not grow for periods of 10-20 years? It’s a pretty long range to swallow when timing own retirement… http://blog.stocktradersalmanac.com/media/image/AIN_0613_20130509_WarInflation500%25Moves.jpg
Hi Alex,
Nope. We don’t market-time our content either. :) This piece was a long time in the works, and we published it when it was ready.
First, I’d like to point out that the graph is based on the Dow Jones, a sample of 30 US companies. I wouldn’t ever use 30 companies to measure the returns to investing . Here’s an article on the Dow I really enjoyed:
http://www.nytimes.com/2012/02/12/magazine/dow-jones-problems.html?pagewanted=all&_r=0
With respect to ‘Retirement Timing’, I think as ever we need to focus on the things we can control – our retirement date, our savings rates, and how much risk we’re taking. The closer we are to a retirement date, the less risk we should take, bearing in mind that we can live up to 40 years in retirement nowadays. To funnel-in on a stable outcome, we should reduce risk over time until we retire, pay attention to our nest-egg, and then in retirement be flexible about our decumulation rate. You can read more here:
https://www.betterment.com/resources/retirement/investment-income-retirement/retirement-income-shouldnt-be-a-guessing-game/
Best,
Dan
Great article, and with very nice animations.
There is nevertheless something very surprising to me because you do not take into account recession periods. I mean, say I have a 70% Stocks/30% Bonds portfolio, I know that during an economic recession, my portfolio will lose some value.
So my question is: why Betterment does not rebalance automatically my portfolio in these specific situations ?
Kind regards,
Stig
Hi Stig,
Betterment will rebalancing you in a tax-aware manner any time your drift reaches past a threshold.
Best,
Dan