Premium Market Analysis

Asset Allocation

EMH Vindication And Why Diversification Is Placebo Effect

Here is a vindication of EMH as far as longer-term investing based on Dow Jones data since 1915. Statistics indicate that diversification has placebo effect on investors and it is not an effective remedy against market swings due to its underperformance.

Let us look at S&P 500 buy and hold since 1960 versus the long-only 50/200 moving average cross timing strategy below. No commissions or dividends are included.

There is a small outperformance of about 12 basis points only for the timing strategy. On a risk-adjusted basis, MAR (CAR/Max. DD) is nearly double, 0.21 for the strategy versus 0.12 for buy and hold. The outperformance can also be seen from Sharpe, 0.61 for the timing strategy versus 0.43 for buy and hold. However, the outperformance is not significant and a switch to a timing strategy with all the associated complications is justified only in the context of lower risk, as returns are about the same.

Next, let us look at buy and hold and the same timing strategy in a longer time period with DJIA data since 1915.

Buy and hold and the timing strategy have the same exactly 5.98% CAR and risk-adjusted performance based on MAR varies little, 0.07 for buy and hold versus 0.10 for the timing strategy.

This is a vindication of the efficient market hypothesis. There is no way to beat the market with a simple timing strategy selected ex-ante.

In the chart of DJIA since 1915 below, we can see that a drawdown of 30% or larger has occurred 19.38% of the time.

The frequency of 19.38% for a drawdown larger than 50% may appear high but notice how most of this was due to the slow recovery from Great Depression.  In fact, in the case of the S&P 500 since 1960, the corresponding frequency is only 0.14%. That is a huge difference.

This low frequency of maximum drawdown larger than 50% in S&P 500 since 1960 does not justify employing timing models or diversification other than for raising the comfort level, i.e., in the form of a placebo. This is unless one expects another Great Depression. In fact, some in financial media have tried to convince investors that another shock of this type is coming. It was probably coming but central banks stepped in and averted it. It is highly unlikely that another prolonged drawdown such as the one shown in the DJIA chart above between 1929 and 1954 will occur again due to central bank intervention. Is this why some also wanted central banks to be abolished? Usually, the same people who were forecasting another Great Depression were also calling for establishment of the Fed. Is this a coincidence? Probably not, they have an agenda and a goal. In short, they are known as permabears.

According to the above limited analysis then, market timing makes sense only in short-term trading, not in longer-term. We have only looked at one timing model without any optimization but we believe that results are not too different for other models. In addition, diversification is a placebo. Below is the (under) performance of the 60/40 portfolio in large caps and total bond since 1972 versus the Vanguard 500 index investor.

Source: Portfolio Visualizer

The benchmark outperforms the passive allocation (rebalanced annually) by about 150 basis points. On a risk-adjusted basis, MAR for buy and hold is 0.22 versus 0.29 for the 60/40 allocation. Again, this is not substantial outperformance on a risk-adjusted bias; it exchanges return for comfort, mainly psychological, and as a result diversification has a placebo effect. In addition, things can get worse if there is a bond and stock crash at the same time, something that has not happened before. In this case, the EMH may surface strong.

The take away here is that market timing in the longer-term may be a waste of time unless a very sophisticate model is used. In the short-term market timing can be rewarding but this is another chapter.

Finally, I will not be writing any more articles for public access with quant analysis for my blog mainly due to plagiarism issues but also because I have to financially support this effort after eight years of doing it. Therefore, I will focus on premium content and make the articles available only to those who subscribe. There is plenty of free material around for non-subscribers. Instructions for subscribing to premium articles are here. Subscribers get a 25% discount for all renewals or upgrades.

If you found this article interesting, I invite you to follow this blog via any of these methods: RSS or Email, or follow us on Twitter

If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY

Charting and backtesting program: Amibroker


Quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.