Would you consider retiring in Singapore

Perfect imperfection


If you knew when a bull market would reach its peak or when a bear market would reach its bottom, investment success would be guaranteed - and a very early retirement. However, our analysis of periods of heavy losses over the past 90 years suggests that it is not necessary to hit the bottom of a bear market exactly. Investors who increase the risk in their portfolio during strong sell-offs can miss the point in time at which the bottom is reached quite significantly and still achieve decent profits. In other words, a contrary approach during a crisis pays off in our opinion, regardless of whether it is possible to perfectly match the absolute low point in the market.

The lessons of the past

There is much to suggest that the market crisis triggered by the coronavirus pandemic can be described as "historical": the type of shock (a pandemic), the pace of the sell-off on the stock markets, the speed with which liquidity dried up on the credit markets, the sudden screeching of the real economy and the scope of stimulus measures. The crisis still lingers and investors around the world are now faced with the difficult question of how - and when - to invest their money. It is important that investors stay diversified and invested for the long term. However, decisions about short-term tactical asset allocation can add value by over- or underweighting risky assets in order to take advantage of short-term relative value opportunities. These tactical decisions should be made in the context of long-term strategic weightings - shifts from 0% to 100% in favor of stocks or bonds should generally be avoided.

It is obvious that it is worth buying risk assets when they are cheap. But in the current environment, choosing the right time is a difficult task. At the center of this dilemma is a very simple question: What is the price if you are too early or too late?

In search of an answer to this question, we examined 17 episodes in which the S&P 500 Index suffered declines of 15% or more, starting with the stock market crash of 1929. These episodes are circled in blue in Figure 1. The last circle marks the current losing phase, which was not part of our analysis.

Since 1928 there have been 18 strong sales

(Fig. 1) The low of March 2020 occupies the shared eighth place

Within the 15% range for the maximum loss.
Within the 15% range for the maximum per loss, after setting the maximum loss during the Great Depression (June 1, 1932) as the lower limit.
Current loss (not part of the analysis).

Past performance is not a reliable indicator of future returns.

As of April 3, 2020.

Source: Calculations by T. Rowe Price based on data from FactSet Research Systems Inc. All rights reserved.
In order to identify troughs or severe loss events, we calculated the losses compared to a previous high (an absolute high) from price data for the S&P 500 for the period from January 3, 1928 to April 3, 2020 and then determined for each high the corresponding days with the highest loss. Then we set a threshold of 15% for the maximum loss to find the days with larger losses in the past. To capture the immediate recessions that followed the Great Depression (February 27, 1933, March 14, 1935, April 28, 1942, and June 13, 1949), we ran the Major Loss Identification process a second time. In doing so, however, we applied the date June 1, 1932 to this data for the new lower limit and starting point, so that the “new highs” and their corresponding losses could be identified.

We calculated the average returns for a hypothetical investor trying to hit the bottom of each of these 17 periods of loss. We assumed that the investor tactically switches from an allocation of 60% stocks and 40% bonds to 70% stocks and 30% bonds if he believes the market has bottomed out. Our analysis showed that the average alpha would have been quite high if the investor had hit the bottom of the market perfectly: One year after switching 10% of the portfolio from bonds to stocks, the hypothetical investor would have achieved an alpha of around 500bps on average .

The astonishing thing is that the investor could have increased the stock position at any point in time within a period of three months before to three months after the absolute low and still would have recorded an increase in value a year later. Even an increase in the allocation to shares six months before or after would have paid off in the end. In other words: even if investors clearly miss the low point, it can still pay off for them to buy more shares in a market crisis.

- Sébastien Page, Head of Global Multi-Asset

Figure 2 lists the results in more detail. So, if we had increased the equity allocation by 10% a month before the bottom, the average twelve-month return would have been 274bps higher than the 60/40 benchmark. If the stock weight had been increased a month after the bottom, the result would have been similar - actually slightly better, suggesting that it is better to trade late rather than early. The most important finding, however, is that increasing the equity component in most of the 17 major loss periods since 1928 would have resulted in higher returns. On the basis of the data for those 90 years, we can therefore say: In most cases, increasing stocks after a strong sell-off on the markets was the right decision.

To validate these results, we ran the analysis again, but this time, instead of measuring the period from the bottom, we measured the average return versus the bottom. The results were similar: even if an investor had missed the bottom by 10% to 15%, whichever side, they would have gained in value in the 17 losing phases a year later, with an average alpha of more than 200 bp.

If equity positions were increased in the event of a sell-off, the returns were higher

(Fig. 2) The effect of increasing the equity position by 10% in a portfolio of 60% stocks and 40% bonds

Past performance is not a reliable indicator of future returns.
As of January 31, 2020.

Sources: Bloomberg Finance L.P., Morningstar (see Additional Information) and U.S. Treasury / Haver Analytics.
This chart is for illustrative purposes only, does not represent an actual investment, and does not reflect the costs or fees of any portfolio. Actual investment results may differ. The stock data is based on the daily total return data in the S&P 500 Index for the period January 3, 1928 through January 31, 2020. The bond data is based on the initial period January 3, 1928 through December 29, 1961 Estimates of daily Ibbotson interpolated returns and total returns on 5-year US Treasuries for the period January 2, 1962 to January 31, 2020, if daily data is available. The results in the table were derived from the daily reweights and indicate the daily cumulative returns. While we're examining all 17 events, some measures in the cells in the chart have excluded the December 2018 event because no data is yet available. The December 24, 2018 event is the most recent event we are investigating. Due to limited data availability, we do not have forward-looking metrics for certain time periods and in these cases we do not include these cells in the calculation of the averages for the summary tables. The cells containing available data for the December 24, 2018 event are: Expected Returns in 1 Month, 3 Months, and 6 Months; expected return in 12 months, except for purchases 3 and 6 months after bottom of the valley; expected return in 18 months only for purchases 6 months before the bottom. The return differential is the difference between the cumulative returns of the hypothetical 70/30 portfolio and the hypothetical 60/40 benchmark, as the average of all 17 past events. The hit rate is the percentage of cases where the 70/30 portfolio outperformed the 60/40 benchmark.

We also measured the effect of daily rebalances on the portfolio; to do this, we compared the returns with those if the rebalancing was done monthly or not at all. We found that rebalancing would have had very little impact on returns.

Investing in the opposite direction can be a clever strategy

The "current" low point in the wake of the coronavirus crisis was reached on March 23, when the S&P 500 index closed 33% below its all-time high of February 19. In the 18 days that followed, share prices recovered more than 25% from this low. This recovery took place significantly faster than in the 17 loss phases of our study, in which they only recovered by 20% on average after 65 days.

- David Eiswert, Portfolio Manager, International Equities

There is currently no clear indication whether the last sell-off on March 23rd hit the bottom or whether this mark will be retested in the next few months. However, our analyzes show that increasing the share allocation during a strong sell-off in the past always paid off in the long term, regardless of how good and precise the timing was. As paradoxical as this may seem at times, the past has shown that increasing the risk in the portfolio can be a worthwhile investment strategy even if it is not certain whether the markets have really bottomed out.

Additional information

© 2020 Morningstar, Inc. All rights reserved. The information contained in this document: (1) is the property of Morningstar and / or its content providers, (2) may not be copied or distributed, and (3) its accuracy, completeness, and timeliness is not guaranteed. Neither Morningstar nor its content providers will be liable for any damage or loss arising out of the use of this information. Past performance is no guarantee of future results.