As of this writing, August 2019, the US and global equity markets have made substantial gains and hit all-time highs year-to-date. However, the yield curve inversion earlier this year and other leading indicators are flashing warning signals of possible recession within the next few months to a year. Based on these indicators, the Fed recently cut rates. We know that recessions are always accompanied by equity drawdowns. Is now the time to sell?
Let’s look at some data. After making new all-time highs, do markets tend to revert to the mean or continue to grow? From Dimensional:
Average Annualized Returns After New Market Highs
S&P 500, January 1926–December 2018
In US dollars. Past performance is no guarantee of future results. New market highs are defined as months ending with the market above all previous levels for the sample period. Annualized compound returns are computed for the relevant time periods subsequent to new market highs and averaged across all new market high observations. There were 1,115 observation months in the sample. January 1990–present: S&P 500 Total Returns Index. S&P data © 2019 S&P Dow Jones Indices LLC, a division of S&P Global. All rights reserved. January 1926–December 1989; S&P 500 Total Return Index, Stocks, Bonds, Bills and Inflation Yearbook™, Ibbotson Associates, Chicago. For illustrative purposes only. Index is not available for direct investment; therefore, its performance does not reflect the expenses associated with the management of an actual portfolio. There is always a risk that an investor may lose money.
This chart strongly suggests that selling following all-time highs may not be the best strategy or timing tool.
Can we glean any useful market timing information from the Fed rate cut? Data is mixed. From DailyFX.com:
Although the average and median returns are all positive for every time period studied here, there is a great deal of variation. Most troubling, the 2 recent Fed rate cut cycles show the only negative returns, and they each exhibit a very different pattern from the older data. Has something changed over more recent history? Or will our current cycle revert to previous patterns?
Dimensional recently studied the performance of actively managed mutual funds and found that even these teams of money managers are not beating the market: over the last 20 years, 77% of equity funds and 92% of fixed income funds failed to survive and outperform their benchmarks after costs. Although market timing is not likely the entire explanation for this, it is certainly a component.
When considering technical or quantitative analysis methods, we can look at many rules-based ETFs. ETFDB.com lists 83 ETFs as momentum- or quant-based. Of these, only 19 (23%) lag their Yahoo Finance category returns YTD. However, if we look at one-year returns, which include the drawdown in December 2018, a whopping 73 of 83 ETFs (88%) are underperforming their category (data from Yahoo finance).
What about valuations, such as the Shiller CAPE/10 or other similar growth/value metrics? John Hussman has studied these extensively. His current hypothesis is that valuation metrics cannot be used as a timing tool. He is studying other market metrics to combine with value, hoping to find the market timing holy grail. Here is the latest market commentary which details these thoughts: https://www.hussmanfunds.com/comment/observations/obs190714/
Once you successfully sell at or near a market top, what are the indicators that it is time to buy back into the equity market (presumably at a lower level)? It’s important to remember that market timing success requires that the timing is correct on two successive transactions, not just one. It is not safe to assume that the market low point is somehow identified by just the reverse, or inverse, of whatever timing tool is used at the top. In other words, your model needs to be proven at both extremes to be successful.
For many advisors and clients, the simplest and most time-proven method of success in owning equities is simply to hold through the downturns, small and large. Regardless of any other metric, more time in the market is associated with a greater likelihood of positive returns. Again, from Dimensional:
Frequency of Positive Returns in the S&P 500 Index
Overlapping Periods: 1926–2018
In US dollars. From January 1926–December 2018, there are 997 overlapping 10-year periods, 1,057 overlapping 5-year periods, and 1,105 overlapping 1-year periods. The first period starts in January 1926, the second period starts in February 1926, the third in March 1926, and so on. S&P data © S&P Dow Jones Indices LLC, a division of S&P Global. Indices are not available for direct investment. Index returns are not representative of actual portfolios and do not reflect costs and fees associated with an actual investment. Past performance is no guarantee of future results. Actual returns may be lower.