Buy and hold vs. market timing is a widely debated theme in the stock market. There are numerous articles on the internet arguing the futility of timing long-term investments. We decided to do some tests ourselves by removing just a few observations from the datasets to see what happens to the long-term compounding:
Buy and hold vs. market timing shows that you increase the risk of getting mediocre returns if you get it wrong. In general, market timing is a pointless exercise if you are a long-term buy and hold investor. Removing just a few of the best and worst days in the price series changes the end result dramatically.
Why is wrong market timing influencing the end result so much? Because of the compounding effect. This is what Warren Buffett calls Snowballing. If you get it right, you become a genius. If you get it wrong, you risk looking like a fool.
Is compunding the eight wonder?
Albert Einstein is famous for saying that compounding is the eighth wonder. But is he correct?
Yes, to a certain extent, he is correct. But Mark Spitznagel made some interesting calculations in his latest book called Safe Haven investing.
Spitznagel says multiplicative compounding is the most destructive force in the universe. In real life, the problem is that you can only traverse one path, not the average. If you don’t pick the right path, you might ruin your compounding ability for years and decades to come. If you get it wrong, the compounding effect is not to your advantage.
Recently, we showed why arithmetic and geometric averages differ in trading and investing by using Monte Carlo simulation. We recommend reading that article if you don’t understand Mark Spitznagel’s reasoning.
Is buy and hold a good strategy? Better than market timing?
If you want to participate in society’s wealth creation, you have to invest in stocks. It has proved to work very well for over a century although some countries went bust along the way, like China, Germany, and Russia, for example (and some others).
But even if your country does not default, you can end up with mediocre returns (even penniless) if you are unlucky with the sequence of returns. For example, those who invested in 2000 didn’t have any returns for over a decade. Currently, many young investors most likely have never experienced a real bear market.
The best advice is probably to invest in stocks and forget about it. Women are better investors than men because they do it simply: they save, invest, and forget about it. They are not trying to be smart and thus beat the majority of those who do market timing.
You have no control over the sequence of returns, but you can control your own behavior:
Market timing: trading vs. investing
Keep in mind that this article doesn’t discuss trading. Trading is something completely different than buy and hold investing. Trading involves exploring and finding market inefficiencies and market edges in order to turn over your capital frequently, Trading is labor intensive – far away from sitting on your ass doing nothing.
(As a humble reminder, we put up our link to our monthly Trading Edge, which we send to our paying subscribers):
Buy and hold vs. market timing: First test
We downloaded daily bars for the cash index of the S&P 500 back to 1960. The cash index doesn’t include reinvested dividends, and thus the CAGR is slightly lower than if you, for example, had invested in a total return ETF or fund (SPY was the first one in 1993).
We did two backtests:
One backtest removed the ten best days of the complete dataset (replaced with zero performance that day), and the other backtest removed the ten worst days of the dataset. Remember that this is ten days of 15 577 days – 0.064%.
Let’s first check the annual result of buy and hold over the whole period:
10 000 invested in 1960 was worth 782 000 at the end of October 2021.
If we remove the ten best days, the value of the investment drops to 350 000. If we do the opposite and remove the ten worst days, the portfolio value increases to 2.24 million.
Quite a difference!
The annual returns, the CAGR, are like this:
- Buy and hold: 7.4% (no dividend reinvestment)
- Remove the ten best days: 6%
- Remove the ten worst days: 9.3%
Even small changes in the annual returns lead to wildly different results in your retirement fund. In the battle between buy and hold vs. market timing this is an advantage to buy and hold.
If we look at the graph, the differences are even starker:
The graph above uses a linear scale and shows buy and hold (blue line), missing the ten best days (red line), and avoiding the ten worst days (yellow line).
What happens if we switch to logarithmic charting? (Read here for an explanation of linear vs. logarithmic charts and scale.) A logarithmic scale shows the percentage moves and is much more suitable for charting the longer the time horizon:
Buy and hold vs. market timing: Second test
Let’s make a new test: we remove the best and worst day over the whole period (one day out of 15 577 – 0.0064% of the dataset). The worst day was the 19th of October 1987, and the best was the 13th of October 2008 (in the middle of the financial crisis). We have provided a list of the best and worst days further down.
When we use a logarithmic scale, we get the following equity curve:
The CAGR is like this:
- Buy and hold: 7.4% (no dividend reinvestment)
- Remove the single best day: 7.2%
- Remove the single worst day: 7.8%
Why do the results differ so much?
When you remove the worst and best days, you start compounding at a higher and lower plateau. This determines your future retirement fund a lot.
This is why Mark Spitznagel is correct in saying compounding might be the most destructive force in the universe. Most private investors perform poorly in the markets, and the main reason is that they get it wrong. And when they get it wrong, compounding has a detrimental effect.
Warren Buffett is not a relevant benchmark. He has admitted that he was lucky when he started investing: he started in the 1950s when the multiples were extremely low after a whole generation was still pessimistic after the Depression. He was lucky with his sequence of returns!
We are not saying Buffett is a poor investor. He is not, he is most likely very intelligent and rational. But it’s impossible to take out luck and randomness from any performance.
The ten best days in the S&P 500 1960 – 2021
The table below contains the dates of the ten best days from 1960 until October 2021:
| 3/10/2009 | 6.36% |
| 11/24/2008 | 6.47 % |
| 11/13/2008 | 6.92 % |
| 4/6/2020 | 7.03 % |
| 3/23/2009 | 7.07 % |
| 10/21/1987 | 9.09 % |
| 3/13/2020 | 9.28 % |
| 3/24/2020 | 9.38 % |
| 10/28/2008 | 10.78 % |
| 10/13/2008 | 11.58 % |
As you can see, six of the ten best days happened in the middle of the financial crisis! If you want to understand why, please read our article about the anatomy of a bear market (2000 -2003). In times of crisis and uncertainty, the markets tend to overreact both up and down. This is why it makes it so extremely difficult to time the market.
The ten worst days in the S&P 500 1960 – 2021
To illustrate the difference buy and hold vs market timing we do another backtest. We remove the worst days from our time series.
The table below shows the dates of the ten worst days:
| 10/19/1987 | -20.46 % |
| 3/16/2020 | -11.98 % |
| 3/12/2020 | -9.51 % |
| 10/15/2008 | -9.03 % |
| 12/1/2008 | -8.92 % |
| 9/29/2008 | -8.80 % |
| 10/26/1987 | -8.27 % |
| 10/9/2008 | -7.61 % |
| 3/9/2020 | -7.59 % |
| 10/27/1997 | -6.86 % |
Buy and hold vs market timing: conclusion
Our very simple tests of buy and hold vs. market timing show the dramatic consequences of long-term compounding by leaving out 0.064% or less of the data in our price series. Admittedly, we removed the most significant days, but it shows how it influences the long-term results.
