Editor's note: This is the last of five columns in which Vern Hayden discusses current mutual fund investment strategies. In the first four installments, he looked at funds with multiple star managers, funds of funds, asset allocation and risk management. Today he examines timing strategies.
Sitting in an open Range Rover, the driver maneuvered around some very tall bushes and almost bumped into the hind legs of a very big elephant. As we came to an abrupt stop, the driver whispered, "Well, timing is everything!" We were in the southern part of
Kruger National Park
in South Africa. It was 1983. The mind works in strange ways because, as the driver said that, I flashed to the image of a coffee cup I had back home. Given to me by a market timer, the cup bore the very same motto: "Timing is everything!"
Is timing everything, and does timing the market work? I have boldly stated on several occasions that it doesn't. Pure timing means there are instances when you have all your money in the market and instances when it's all out and sitting in a money market fund. The premise for this strategy is extremely seductive. It simply says that as the market goes up, you get in, and as the market starts going down, you get out. Your investing pattern will look like a series of stair steps going all the way to the stars.
The reason I have been so critical of timing strategies is because, from the mid-1970s to the early 1980s, I experimented with three different timing services. Each of them suffered from the same problems. All three used mutual funds. Here were some of the problems:
- The market would quickly turn on them, nullifying the effect of their most recent move. In other words, they would get into the market and then almost immediately get a sell signal to get out. This whipsawing effect disables the strategy.
Each timer had a proprietary system of technical indicators. These indicators were supposed to be "fail-safe" because they had been back-tested over a number of years. The concept of back-testing always bothers me because you can do a lot of strange things with history. There's nothing like testing in real time. Sooner or later, the timers I used made big mistakes. Making big mistakes doesn't make you popular with clients.
Mistakes are hard to explain to clients because being a timer connotes some element of superiority to the market. If you are smarter than the market, how can you make mistakes about its behavior? When you create a higher standard than the market, you're going to have further to fall. That means poorer performance and fewer clients.
Every timing change would create a blizzard of paperwork. Confirmations of transactions from the funds would fill the clients' mailboxes -- a small irritant at first, but a significant one after a while.
Timing is expensive. The minimum fee for the timing service was 2% per year on the total value of the account. This was in addition to the mutual funds' annual fees. The timing service had to make 3% to 5% a year just to break even. I'm not one of those people who knows the cost of everything and the value of nothing. However, when I am told a service brings added value to my investment results, I hope to see those results, and I didn't.
All timers purport to lower the risk of investing. Every timer seems to have a much lower "beta" than the market. I think most timers probably are out of bear markets and miss most of the corrections. That is when they are supposed to have their best performance. The one timer I was "testing" in 1987 was half in the market and half out. It is almost impossible to explain to someone how a timer can miss a crash. It is at that point that a timer's credibility crashes. This timer proceeded to stay out of the stock market for the next three years, totally missing a rebound. What kind of beta is that? By the middle of 1988, I decided to end my experimentation with all timers.
The paradox of the timing strategies I tested was that, in every case, a well-managed growth mutual fund far outperformed the timing service.
Aside from my personal experiences with timing, I am sure that you would appreciate a more current and objective assessment of timing. To help overcome my bias, I called the man who probably knows the most about timing services, Steve Shellans.
Shellans actually tracks about 100 timers. He is in Portland, Ore., and since 1981 has tracked timers. In 1986, he started publishing a newsletter entitled
(1-800-615-6664). In his newsletter, he gives actual, real-time-audited performance numbers on the timers. I subscribe to his newsletter, so I have seen these numbers for the last few years.
I asked Shellans to define timing. He stated that while a definition of timing is a "little bit slippery," the following would work: "Timing is done by managers who shift substantial portions of assets from asset class to asset class. Some switch every few days, and some switch as little as twice a year."
To give you an idea of performance, here are the numbers from the March/April 1999 issue of Shellans' newsletter based on year-end 1998. These numbers are net of fees and annualized.
Shellans referred me to a timer with an outstanding track record. When I spoke with the timer, he explained that his biggest problem was the capacity to handle all the money coming in, given his modest staff. This timer only used international funds, and his returns have averaged 36.4% for the last five years and 40.3% for the last three years, net of fees. These are audited performance results. It is tough to argue with those results! Obviously, his method of timing is working for him. He has no doubt about his ability to perform indefinitely. Sounds like an ad for
Shellans also feels that toward the end of 1999, there may be a number of things coming together to make that next bear market happen. Y2K problems, depression of corporate earnings and overvaluation are just a few of the factors that would start a correction. If so, that would put timers back in the limelight.
Does timing work? Most timers don't seem to work. In Shellans' universe of timers, it seems the ideal system would be one that tells you when to change timers, since no one timer has been right for 20 years.
I think there is a broader definition of timing. This definition would include the timing decisions a mutual fund manager makes in buying and selling stocks. It would also include the kind of timing decision that I made when I got out of European funds last year. I'll have more on this in my wrap-up on all the strategies next week. In the meantime, please send me an
email telling me what has or hasn't worked for you.
Vern Hayden is a certified financial planner with American Planning Group in Westport, Conn. His column is not a recommendation to buy or sell stocks or to solicit transactions or clients. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks or funds. While he cannot provide investment advice or recommendations, Hayden welcomes your feedback at