For market timing to work successfully, at least several essential things must be present.

Among them:

In the following discussion, I want to look at timing applied to two mutual funds over a 10-year period ending in September 1998. The timing in this study is what we call fund timing, not market timing. That means a timing system (in this case a 75-day moving average with a 1 percent envelope, for the technically inclined) is based only on price movements of one mutual fund, and it is meant to apply only to that fund. Market timing, by contrast, is based on an entire asset class and is suitable for use with multiple funds.

For the sake of this discussion, let’s assume that our timing system is suitable, or that it’s the best one we can devise. Therefore, "Get a new timing system" is not going to be a valid response to disappointments.

What we will see is that even with perfect discipline and a good timing system, it can be very discouraging to follow it.

The two funds we’ll study are Invesco Pacific Basin, which invests in stocks of Pacific Rim countries, and Invesco Dynamics, which invests in U.S. small-cap companies.

Invesco Pacific Basin Fund

Let’s start our study with Invesco Pacific Basin. Figure 1 shows two views of the 10-year performance of this fund, from October 1, 1988, to September 30, 1998. The bottom graph shows the fund on a buy-and-hold basis. If you bought shares in the fund on October 1, 1988, and held on for 10 years, your investment would have tracked the line in this graph.

Invesco Pacific Basin Fund varied widely in price from October 1988 through September 1998, as shown in the bottom graph. Arrows indicate buy (pointing up) and sell (pointing down) signals generated by this fund's price movements. The top graph shows the effect when these timing signals were followed. The scales on the sides indicate actual dollar prices in the bottom graph and implied or effective prices in the top graph.

In the bottom graph are 50 arrows. Half of them point upward, representing buy signals generated by this timing system. The other half, pointing downward, represent sell signals.

If you had timed this fund with the system we used for this study, you would have started in cash and made the first of 25 buys in January 1989. You would have sold 25 times as well, and you would have been out of the fund, in cash, since August 1997.

You could think of each signal pair, a buy signal followed by a sell signal, as a separate interim investment in this fund. In between each interim investment, we assume you parked your money in a money-market fund.

Without timing, a $10,000 investment in this fund "grew" to $7,440 (actually a loss of 25.6 percent) over the 10 years, including reinvestment of dividends and capital gains distributions, making it a lousy investment. With timing, an equal $10,000 investment grew to $15, 911. That’s still not very great for a whole decade, but it’s vastly superior to a 25.6 percent loss.

Results of the timed strategy are shown in the graph at the top. The almost-flat lines represent times when your investment was in cash. When this works as it is supposed to, the top graph shows a straight line with a very gradual upward rise, representing the stability and interest from a money fund, directly over a declining line (in the bottom graph) that shows a drop in the price of the fund. That drop represents the loss from which timing is supposed to protect you.

A fine example of that occurred with Invesco Pacific Basin late in 1997 and through the first three quarters of 1998. A sell signal was generated on August 12, 1997. Shortly after that, the fund itself took a nose dive, losing about half its value by the end of the third quarter of 1998. In that same period, as you can see in the top graph, the timing strategy not only retained its value but rose modestly as it earned money-market interest.

Timing can be tough

Is this just a way for us to advertise the benefit of timing? If you think so, read on. Because we’re going to show you how hard it was to actually time this fund and achieve that superior performance.

But first, think for a moment about this interesting question: Why would you, or anyone, buy this particular fund anyway? Obviously if you knew in advance the fund’s 10-year performance, you probably would have chosen somewhere else to put your money. But there are two likely reasons you might have invested in this fund late in 1988. First, you may have believed in Pacific Rim investments. It was (and still is) a region with enormous financial potential. But you might also have invested in this fund because of its recent past performance. That’s the biggest factor that drives investments: recent past performance. This fund, which was then known as Financial Strategic Pacific Basin, gained 27 percent in 1985, 72 percent in 1986, 9.8 percent in 1987 and 23 percent in 1988 (the last part of which you see on the graph). That track record could be counted on to attract lots of investors – just as the Standard & Poor's 500 Index has attracted so many investors over the past few years.

But many of those new investors subsequently became disillusioned. For the first four years shown on the chart, the fund ended slightly lower than where it started, with some wrenching volatility along the way. Notice that timing, by avoiding the worst of the losing periods, produced a modest gain in those four years. But the timing results for 1992 were mostly a series of losses interspersed with money-market investments. In hindsight, that would have been a good year to simply stay in cash.

The next two years, 1993 and 1994, provided the type of results that investors cherish. If you compare the top and bottom charts, you’ll see there’s not much difference between timing and buy-and-hold. The reason? In a powerful bull market, timing simply can’t add much value.

The "hot money" cools off

But how many of those hopeful 1988 investors were still around in 1993 and 1994 to reap the benefits of this powerful upward run? Probably not many. Within the industry, it’s well known that a lot of capital predictably chases recent performance. This is known as "hot money." And hot money usually doesn’t have much patience for investments that appear to have turned cold, as this fund did from 1989 through 1992.

For about two years, from early 1994 to early 1996, Invesco Pacific Basin made a nice upward move, then lost most of it, only to regain it again and wind up about where it started. Such a pattern should provide a good opportunity for timing. But as you can see in the top graph, timing did not avoid enough of the losses to make a dramatic difference.

Now let’s "zoom in" on part of this period and examine three interim investments dictated by the timing system in 1992 and 1993. The moves you’ll see in Figure 2 happened in this particular fund. But the same patterns could occur with any asset. In fact, what you’re about to see is typical.

Detailed graph of Pacific Basin Fund's performance in 1992 and 1993. Note the lack of any advance clue that the December 1992 buy signal was much more valuable than the previous ones.

The first buy in Figure 2, shown with an up-arrow, occurred in May 1992 at a fund price of $9.05. This signal came near the end of a strong upward move. The fund price soon peaked and headed south, and a sell signal, shown by a down-arrow, ended that interim investment in June at a price of $8.69 (for a loss of 4.0 percent).

The next buy was in October at $8.39, followed by a sell early in December for $8.23 (a loss of 1.9 percent). Imagine your state of confidence at this time if you were an investor. Your fund has been essentially going nowhere for several years, and the last two times you bought shares in this fund, you’ve been left with losses.

So near the end of December, when the price had risen to $8.46, would you have been eager to leap back into the fund when you received another buy signal? Or would you have thrown in the towel by then?

Is the 13th time a charm?

By this point in our study, you had bought 12 times and sold 12 times. Only three of your interim investments had been profitable; the other nine had brought you grief and frustration.

You could not have known it at the time, but it was the 13th buy signal that held the charm. From the chart in Figure 2, it’s easy to see: that buy signal turned out to be the one you’d been awaiting for years. Had you followed the system, over the next 11 months you would have gained 32.6 percent as the fund price rose to $11.22 by the time you reached a sell signal in November 1993.

Remember that you chose this fund initially hoping for dramatic gains. But if you didn’t follow that 13th buy signal, you missed out on the only dramatic gain that this fund delivered during this 10-year period. Only by keeping the faith and following that signal did you get what you came for. Would you have done it?

Now I’d like you to place yourself mentally in the summer of 1997. As you can easily see in Figure 3, your fund has peaked (though you could not have known that at the time). If you had followed all the timing signals to this point, they had added a modest amount of value, though they had put you through a lot of frustrations.

Detailed graph of Pacific Basin Fund's performance in 1997 and 1998. The August 1997 timing signal, if followed, preserved the previous nine years of gains.

Can you imagine yourself having grown weary of following sell signals only to get back in when the next buy signal arrives and the fund price is higher? Can you imagine yourself thinking that the sell signal in August 1997 might be one you could ignore? Like the child who "cried wolf" too many times in the old fairy tale?

You couldn’t have known it then, of course, but that mid-1997 sell signal was "the big one" that you had been waiting for. It came in just the nick of time, before a free-fall plunge that stripped more than half the value from your fund in the next 13 months.

If you followed that sell signal in 1997, you preserved your capital and made some money on your cash. If you did not follow that signal, you very quickly lost the benefit of everything you had done in the previous nine years.

Now here’s a recap of what these charts tell us: If you were timing this fund, over a 10-year period you bought 25 times and you sold 25 times, following 50 individual timing signals. It turned out that almost all the value from timing came from just two signals: the buy at the end of 1992 and the sell in mid-1997. The other 48 buy and sell signals in retrospect were little more than noise.

This is one of the best case studies I’ve ever seen to show that timing works only if you follow it all the time. If you pick and choose which signals to follow, you’re just relying on luck, and you might as well consult tea leaves or the horoscope in your daily newspaper.

Invesco Dynamics Fund

Now let’s take a look at the other fund we mentioned. Figure 4 shows the performance over the same 10-year period of Invesco Dynamics Fund, again with and without timing. This successful U.S. small-cap growth fund’s favorable returns (approximately break-even in the first nine months of this year vs. a 15 percent drop for an index of small-cap funds) and below-average volatility have made it an attractive package for aggressive investors. But that kind of low volatility, which by the way is quite unusual for a small-cap fund, is a challenge for timers.

Invesco Dynamics Fund showed impressive growth from October 1988 through September 1998, as shown in the bottom graph. Arrows indicate buy (pointing up) and sell (pointing down) signals generated by this fund's price movements. The top graph shows the effect when these timing signals were followed. The scales on the sides indicate actual dollar prices in the bottom graph and implied or effective prices in the top graph. A decade of timing signals reduced this fund's performance significantly.

As you can see in the chart, the past 10 years have been very good for this fund. If you look at the top chart, you’ll find a few of those flat-bottom lines indicating periods when fund-timing would have put you on the sidelines. And in every one of those periods, as you will see if you look straight down at the bottom graph, the price of the fund itself declined.

Yet, these two graphs echo the familiar refrain that during a long bull market, timing can’t do much to add value unless there is volatility. Invesco Dynamics simply didn’t have the volatility to let timing move in and out while capturing gains and avoiding losses.

Let’s "zoom in" again (Figure 5) and look at four frustrating interim investments in this fund late in 1997 and early in 1998. Early in November 1997, the fund timing system generated a buy on Invesco Dynamics at a price of $14.02. The fund rose for a few days, then plunged to $13.25 a week later, generating a sell signal on what turned out to be the day after it hit its low point and started back upward.

Detailed graph of Dynamics Fund's performance in late 1997 and early 1998. The first nine timing signals in this period were all counter-productive. But the buy signal on January 30, 1998, unobtrusively marked the start of the fund's most dramatic gains of a decade.

Two weeks later, the next buy signal showed up at $13.53, and the fund experienced two weeks of gains before suddenly plunging in the first half of December. A sell signal ended that investment at $13.38.

The next buy signal came in the final week of 1997 at $13.96, just before a sell signal arrived at $13.03, the very bottom of that drop. Just 10 days later, when the fund was up to $13.78, "buy" flashed onto the screen once again, only to be followed one week later by a sell at $13.47.

Now imagine yourself as an investor who is timing this fund, and you are notified on the last day of January 1998 that there’s another buy signal. The price of the fund is back up to $13.80. The last four times you’ve followed those buy signals, you have lost money.

Can you imagine yourself feeling sort of jerked around by your strategy of timing this fund? Can you imagine deciding you had had enough of this torture? Can you imagine being tempted to just "wait this one out" and not buy again?

If you’d given in to that very understandable temptation, and ignored this buy signal, you might have watched from the sidelines as the fund price soared to $14.50 one week later and to $15.25 by late February. How would you have felt when the fund reached $16.75 in mid-April?

Is the fifth time a charm?

Out of five buy signals in a short span of time, four turned out to be wrong. But the fifth was the one you’d been awaiting for a long time, and it contained the seeds of one of this fund’s biggest moves over the whole decade.

Again, it shows that if you want the benefits of timing, you have to follow every signal, every time. The vast majority of them may bring you frustration. But all it takes is one or two that are right, and it’s all worth it.

The chart of Invesco Dynamics illustrates something I said at the start of this article: Successful timing requires an asset with sufficient volatility to let timing make a difference. That essential element is missing in this chart. Except on the upside (which doesn’t help a timer), Invesco Dynamics rarely strayed far from the 75-day moving average that established the base for this timing system.

This chart also shows how timing reduces risk. In the top graph, every one of those almost-flat bottoms represents time an investor was in cash. During those times, the risk of this fund was reduced to zero. A casual glance at that graph suggests that an investor who used timing was in cash for a total of about two years in this decade. That means that in a very real way, timing reduced the risk of investing in this fund by about 20 percent.

Normally, you pay a price for reducing investment risk, and this period was no exception. If you followed all these timing signals in this fund, a $10,000 initial investment grew to $31,000. Had you invested in the fund without timing over the same decade, $10,000 would have grown to $46,880. Timing reduced both the risk and the return.

Is timing worth it?

In retrospect, of course, you could always decide that you didn’t need the protection of timing, because you made plenty of money without it. But that is not really a valid comment, because you can never know in advance what’s coming down the pike. Timing is a little bit like a fire insurance policy on your home. Even if you don’t have a fire, it may be worth your while to pay the premium instead of take all the risk yourself.

Here’s another way to view the issue of market timing. Look back at the chart of Invesco Pacific Basin, and ask yourself: In how many calendar years did timing add value? It looks to me as if timing added value to Invesco Pacific Basin in 1990, 1997 and 1998. That’s three years out of 10. Yet the final result, as we have seen, was a huge difference at the end of the period. In the case of Invesco Dynamics, timing added a modest amount of value in 1990, 1992, 1994 and 1998, but its overall result was still negative.

Here are two other things I think these graphs show us.

The bottom line as I see it is that if you evaluate every buy and every sell, you are missing the whole point of what timing is supposed to do. Short of pure luck, you’ll never have a "perfect" timing system. Instead, you’ll have a series of frustrating moves in and out of an asset or a market, punctuated occasionally by big moves that will make it all worthwhile. But most of the buy signals and most of the sell signals, like most daily and weekly fluctuations, are little more than "noise." They are annoying. But you have to tolerate them in order to take advantage of the big moves.

Finally, there’s the concept of diversification. Invesco Pacific Basin’s 10-year record is a good reminder that you can easily get into an asset that goes nowhere for years, with or without timing. That’s why it’s imprudent to put all your eggs in one basket.

 

Discover how professional money management can help you.  Get a Free Consultation from a Merriman financial advisor.