Since the end of last year, I told myself I want to be a defensive investor, ie. investing a fixed amount in STI ETF at the end of every month.
However, when the market advanced more than 20% from its low of 1,456.95 (9 March 09) in less than a month, I sold 50% of my investment in STI ETF, thinking such rapid advancement in a bear market was likely to be corrected. After my sale, the market did correct but disappointingly, negligibly little. After which, it proceeded to rally on. *roll eyes*
Benjamin Graham proposed that an investor should only sell his investments when he needs the money or when rebalancing his portfolio. Frankly, I am quite puzzled. Should a defensive investor not make any attempt to time the market AT ALL? Hold on to his investment through the peak in 2007 and stomach its dwindle in 2008? Somehow, I am more inclined to think that one should take some profit off the table every now and then, especially after steep market rally.
Anyway, I thought of a few strategies for defensive investors and did a comparison of their results by back-testing using STI historical data. I was quite surprised with some of the findings.
Assumptions:Assuming we have a defensive investor who has an additional $5,000 available for investment (after setting aside adequate funds for emergency and liquidity needs) at the end of every year. He can choose to invest the money, partially or in full, in equities or save it in the bank.
For equity investment, we assume there is a STI ETF since 1987 and its price is exactly 1/1000 of the STI Index. We further assume his equity investment provides him with a steady dividend yield of 4% and his savings in the bank compound annually at an interest rate of 2%. For simplicity, theses rate stay the same throughout the 21 years from 1987 to 2008.
On the last trading day of every year, the investor reviews his portfolio: investment in equities (mark-to-market) and cash (cash in bank + additional $5,000 + dividend and interest for that year) and execute his investment decision at the closing price. We also assume that transaction fee is negligible.
The Strategies:(1) Hide all money under the pillowThis is not an investment strategy but is included for comparison purpose.
(2) Put all money in the bankSame as (1), this is not an investment strategy but is included for comparison purpose.
(3) Invest $5,000 yearlyThe investor invest $5,000 at the end of every year but keeps his cash dividends in bank.
(4) Invest $5,000 + dividends yearlyIn addition to $5,000 of new investment, the investor reinvest all his dividends, leaving no cash in his portfolio.
(5) Rebalance portfolio: 50% cash: 50% equitiesThe investor maintains a fixed allocation of 50% equities and 50% cash in his portfolio.
This strategy allows the investor to "buy low sell high" as the investment loss (mark-to-market) during bear market will cause the proportion of his equity investment to fall below 50%, thus requires him to invest more in order to maintain the desired allocation. Similarly, during a bull market, his investment gain would inflat the proportion of his equity investment and thereby prompt him to sell in order to maintain the desired allocation.
(6) When market is up by X%, sell X% of investment; when market is down by Y%, invest Y% of cashEg. when the market goes up by 30%, the investor will sell 30% of his investment. However, when the market drops by, say, 45%, he will invest 45% of his cash. Like strategy (5), this allows the investor to "buy low sell high". The faster the market accelerates, the greater reduction of his market exposure will be, therefore reducing the chances of the investor getting caught in a bubble. Conversely, the harder the market falls, the more he invest, at deeply discounted price.
(7) Sell 50% of investment when market is up by >20%, invest 50% of cash when market is down by >20%, otherwise, invest $5,000This is a modification of strategy (6). Intuitively, strategy (6) will not be very efficient if the market fluctuates only mildly every year for a long period of time (imagine buying and selling less than 1% of portfolio, especially when the portfolio is small). Compared to (6), strategy (7) only sells or increase the investment amount when the market moves in a significant way. Otherwise, the investors invest $5,000 dilligently like in strategy (3).
(8) Sell 50% of investment when TA = sell signal; invest 50% of cash when TA = buy signalThis strategy involves simple technical analysis. For this illustration, the indicator is the 200 days moving average. No special reason for choosing this indicator other than it is simple and supposedly more suitable for longer term investment.
Buy signal = price closes above 200MA
Sell signal = price closes below 200MA
As moving averages are trend indicators, this strategy suggests that the investor reduces his market exposure when the market is on a downtrend, so as to avoid further deterioting of his investment and increase his investment to ride on the market's uptrend. However, as moving averages are lagging indicators, that means the investors are reacting to the market: selling when market is already moving down and buying when market is already moving up.
[The following section is updated on 17 April 2009]
The Results:Figure 1: Different results during bull and bear market
Table 1: Portfolio sizes under different strategies over the years
The numbers under the columns entitled (1) - (8) in Table 1 are total values of all assets (equities + cash) achieved using the respective strategies at the given year. The largest porfolio size in every year is highlighted in blue. (Not to worry about the figures, the pattern of the blue highlights are more important.)
My Analysis
If the investor hides all his money under the pillow, he would have saved a total of $110,000.00 after the 21 years. However, by putting all those money in a savings account which gives an annual interest rate of 2%, his wealth would have grown to $136,494.92 at the end of 2008, an increase of 24.1%
Nevertheless, he will end up an even wealthier man if he adopts any of the investment strategies, ie. (3) to (8). At the end of 2008, (8) yields the best results with a portfolio size of $228,168.19. This is 107.4% more than the ending portfolio under (1) and 67.2% over that of (2). Strategy (4) is next, with a portfolio size of $194.067.70, ie. 76.4% and 42.2% above that of (1) and (2) respectively.
From the Table 1, the pattern of the blue highlights seems to indicate that (4) is the dominant strategy most of the time. If we were to end the above exercise in 2007, (4) is clearly the most superior strategy, outperforming (1) by 229.9%.
Apparently, (4) is most favoured by the bull because it is the most aggressive strategy, requiring the investor to be fully invested at all times. With 100% market exposure, the strategy will flourish / perish with the market. Table 1, also seems to suggest that (8) may be the best insurance against market downturn.
To gain more insight, I ran the numbers at different starting points (with minimum investment horizon of 10 years) to see if the timing of commencement of investment affects the results. Interestingly, similar pattern prevails, suggesting that no matter when (bull or bear) the investor starts investing, (4) will usually result in the highest amount of assets, i.e. until the bear strikes.
However this time, although (8) still ended up with the best portfolio in 2008, the blue highlights are distributed across (6), (7) and (8) during market downturns. So, maybe (8) is not be the best insurance against the loss of wealth after all. In fact, if we were to compare the annual % change in wealth (as oppose to the absolute amount), (6) seems to offer better insurance (least %decrease) more often than (8).
(to be continued... see part 2)
