Those familiar with Shakespeare should recognize that as coming from a scene in Macbeth in which three witches meet in a dark cave to do the kinds of things that witches are usually assumed to do when they meet in dark caves. Those who aren't up on Shakespeare might, however, mistake this as something from a treatise on market timing; perhaps a ritual that precedes the shuffling of a deck of tarot cards. It would be an easy mistake to make. Market timing is often assumed to have more in common with sorcery than serious analysis.
But before we chuckle and move on, let's pause and try to take a more dispassionate look at the topic. We know it's hard (if it were easy, nobody would have lost money in 2008 and nobody would have gotten caught up in the market's current downturn), but is it really impossible?
We're going to explore this by assuming we'll invest in the S&P 500 SPDR (SPY) and testing a couple of timing strategies to see how effective they are in telling us when we should move to the sidelines. Figure 1 helps us establish a benchmark. It shows a 3/30/01 - 5/25/10 StockScreen123.com backtest of a buy-and-hold strategy for SPY.
Figure 1
(Notice that the red line, which represents our SPY strategy, is above the blue line, which depicts the S&P 500. That reflects slippage — the gap between the ETF's goal of matching the S&P 500 and what it actually achieved. That's an interesting topic, but one for another day.)
Now, we'll add about as simple a market timing strategy as one can devise. We'll be in SPY only when the 50-day moving average of the S&P 500 is above its 200-day moving average. At other times, we'll be on the sidelines earning what we assume will be a zero return. Figure 2 shows the results of the backtest, assuming we review our timing model and act, if necessary, once per week.
Figure 2
Translated to numbers, $1,000 invested in the strategy grew to $1,824 versus $937 for the S&P 500, and $1,099 for being in SPY at all times. The pre-2008 high for the timing strategy was $1,606 (reached on 9/29/07), versus $1,352 for the index and $1,501 for the permanent position in SPY.
That's interesting. One of the easiest timing strategies imaginable, one that can be implemented in less then a minute by anybody with access to price charts on free web portals, could have kept us on the sidelines for most of two big bear markets of the 2000s. It wasn't perfect. Using 50- and 200-day moving averages makes for slow signals. But even with that, we still could have been spared a lot of financial pain.
Notice, though, that it has not yet signaled the May 2010 drop. If the weakness lasts long enough, the timing strategy will catch up, but the swiftness of the early slump has been noteworthy enough to make us wonder if there is a way to generate a quicker sell signal. Obviously, use of shorter moving averages would help. The problem, there, would be too many false signals at other times.
Figure 3 shows the results of a conditional moving average timing signal I created on StockScreen123. It normally compares the 50- and 200-day moving averages, but switches to a protocol that takes us out of the market if the 5-day S&P 500 moving average drops below its 20-day moving average at times when the CBOE volatility index, the VIX, is 30 or higher.
Figure 3
It's hard to see, but if you look closely, you'll notice a little horizontal red line sticking out on the right, indicating that this strategy did recently signal a down market (specifically, on 5/1/10). The long-term performance of this strategy is not significantly different from the basic 50-200 day moving average comparisons — $1,000 in the conditional strategy grew to $1,803 as of 9/29/07 and $2,062 as of 5/25/10, but it will be interesting to continue watching in the weeks and months ahead, to see if it's worthwhile to continue to use VIX to help us decide if we should work with longer- or shorter-term moving averages.
Obviously, I do not now have all the answers. More study is needed. But it does, at least, seem apparent that I should be working with screening and backtesting, as opposed to books on tarot or cave-based cauldron cookery.
No comments:
Post a Comment