ULTRA Financial Systems

Market timing is simply the act of selling a portion of your stock market positions to a risk-free interest bearing account when odds favor a significant stock market decline. There are very few advisors that do not use some degree of market timing.

Individuals use timing. Most just don't realize it. Every investment has to be bought at some point and sold at some point. The so-called "buy-and-hold" investors end up using a very bad timing strategy based purely on emotions.
Mutual funds may hold 10-15% cash positions sometimes. But they can't risk under performing their competitors by being only partially invested. (They would rather lose money for their clients if everybody else were losing money, than under perform in a bull market.)
Conservative advisors recommend stock positions of 0% to 100% based on their market perceptions.
Aggressive advisors will recommend positions from 100% short to 200% long using margin.

It's easier to accept small losses. Taking periodic small losses will not cause you to give up on the stock market. On the other hand, a 35% drawdown on your account will definitely have you doubting your strategy.
Much of the time you are invested in risk-free cash. Your money is safely earning interest and you get a break from market induced anxiety.
You have a strategy. Without a strategy, you'll always be thinking, "Shouldn't I educate myself and get a strategy?" This question will especially haunt you when the market has turned against you. The temptation to sell everything until you HAVE a strategy will be enormous and irresistible.
Performance. Many timing strategies have outperformed buy-and-hold.
Sophistication. The education you receive from developing a market timing strategy is valuable. It will pay off sometime down the road even if it temporarily isn't working out as well as you'd like.

Most investors use random timing methods. To make their timing decisions they use

This is the type of timing that never works over the long term and gives market timing a bad name. It's driven by emotion and greed and always results in buying high and selling low.

The analysts on TV and Magazines may very well know what they are talking about. But, analysts can change their opinions quickly. For example, you see a well respected expert on CNBC say that he see the market going much higher for the rest of the year. So, you feel confident about your existing positions and may even add to them. However, after a couple of weeks this advisor could easily reverse his opinion. You'd probably never even know. Your emotions will force you to follow another bullish advisor's opinion, as you desperately want the market to go higher. Following this type of random advice is a very bad strategy.

Using unproven methods for market timing is a close relative to random timing methods.

Mechanical timing means that you develop a strategy for timing purchases and sales of investments that does not require any type of expert analysis. A mechanical timing system simply generates buy and sell signals that you follow without question. Mechanical timing methods can be backtested against historical data to ensure that they are historically profitable. Of course this is no guarantee of future performance.

These are the types of timing systems that are contained in ULTRA products. If a timing theory can't be represented in some mathematical form that can be back-tested by a computer, there's really no way to know if it is REALLY historically profitable. For example, you could use a stack of historical charts to test manually drawn trendlines, but no matter how hard you try to avoid it; the results will be skewed based on what you know about historical market action.

An expert makes decisions based on everything that he knows. As you can probably imagine becoming an expert is very difficult. An expert can do very well timing the market versus a mechanical method because:

Many experts utilize mechanical timing methods to help them make decisions. They will then sometimes override the mechanical methods based on knowledge that can't be represented with computer modeling.

Experts will also often allow mechanical methods to manage part of their capital thereby taking pressure off themselves to make good decisions with their entire nestegg. It is very comforting to know watch your mechanical systems make money while you are bearish on the market and can't find a good reason to safely take positions with your discretionary capital.

Mechanical timing methods operate on stock market closing data. This data is only available AFTER the market closes. You can't act on buy or sell signal based on this data until the next trading day. With mutual funds, when you purchase you get the closing price of the fund at end of the NEXT market close.

Most stock market timing advocates tabulate their historical results assuming that they can get the closing price on the SAME DAY as the buy or sell signal occurs. We call this SAME DAY trading. Tabulating results based on SAME DAY trading is usually misleading.

In some cases, SAME DAY trading can be approximated by using PRE-CLOSE stock market statistics. For example, at Rydex Mutual Funds you have until 2:45 PM CST to enter your trade to get the price at the end of that day. It is possible to use 2:40 PM CST stock market data and then enter your trade before the 2:45 PM deadline thereby approximating SAME-DAY trading..

NEXT DAY trading means that you use the actual closing stock market statistics to generate your buy and sell signals. Then, you just casually enter your trade sometime before the market closes the next day.

You need to be careful when deciding on a market timing strategy, as there are many pitfalls. Here are some that you should be sure to understand.

Even excellent systems go through periods where they under perform the market. It's a fact. If you expect to beat the market every single month of the year using timing, get ready for disappointment. Outperforming the market on a quarterly or yearly basis is much more realistic. Here's a very common example, a system stays invested during an 8% drop of about two weeks. Just as an inexperienced user will give up on the system and sell out, the market turns and rallies 25%. Or, a system misses an important bottom and the market rallies 8%. Has this system stopped working? Who knows? Many different scenarios could emerge. This could be a bear market rally. Or, the system may enter later just before a more important market rise. And so on. The bottom line is to be slow to pick systems and equally slow to give up on them.
Some systems are just theories and never worked. In years of designing and testing systems, I'm amazed at the large number that do not work when tested. Some common indicators and systems are of absolutely no value whatsoever. A good rule of thumb is to assume that an indicator or system is worthless unless you have proof otherwise.
Depending too heavily on a single timing system. Timing systems are not perfect. Sometimes they temporarily under-perform the market. Some simply stop working. This is why it is imperative that you do not rely too heavily on a single system. Instead, you should pick a number of good systems and devise a strategy based on the set of systems that move you incrementally in and out of the market.
Assuming the "best" system is the one that has the highest historical return. Your search should be for the system that has, "The best probability of performing well in real-time within your risk comfort zone.." The "Best" system also depends on the individual, SAME DAY vs. NEXT DAY, Frequency of trades, etc.
Assuming a system that performs well with SAME DAY trading will also perform well with NEXT DAY trading.
Curve Fitting. This is adding complexity to a system in order to produce better and better historical results. For example, A simple system that is profitable but only 60% of its trades are winners. You notice that many of the "bad" signals occur near the beginning of the month so you filter those out by adding complexity that brings your accuracy to 70%. As you add complexity you can continue to improve your historical record to 90% accurate. However, the complicated system is much less likely to reproduce the 90% accuracy than the simple system was to reproduce the 60% accuracy.
Optimizing over an entire set of data. Timing systems should be developed using a period of data and then tested over an out-of-sample period to get a pseudo-real-time test of the system.
Testing a system over too short a historical period. It is very easy to design systems that perform well over short periods in history. However, most will fall apart in out-of-sample periods and in real-time. The longer a system performs successfully in historical-testing the greater the odds of real-time success.

During proper construction of a timing system, not all of the historical data is used during the construction process. Some of the historical data is reserved for testing the system in a pseudo-real-time test called out-of-sample testing. Out of sample testing is important to insure against excessive curve fitting.

For example, assuming you have 55 years of historical data, you may reserve the oldest and newest 10 years of data for an out-of-sample test of the completed system. The only problem with this method is that you may have personal knowledge of the latest 10 years and you may use it in the construction process without realizing it.

You can avoid this possibility by using the latest 15 years of data to construct the system and then use the earlier 40 years to run an out-of-sample test. This allows you to construct your system based on recent market action and use the earlier data to make sure the system doesn't bankrupt itself in a long out-of-sample test during different market conditions.

In any case, the longer the out-of-sample test, the better. Be wary of any system that is not tested against out-of-sample data.

"Whatever annual gain you are seeking with a strategy, you will almost certainly experience an equal drawdown in real-time."

Can one substantially beat the market with far less risk? DEFINITELY. The market historically returns about 9% and often takes 40% drawdowns. That is what you'll eventually achieve if you stick with buy/hold over the long term.

However, it's my opinion that anybody armed with a good strategy can earn 15% annually from the stock market while keeping periodic drawdowns around 20%. This is great performance and should be your realistic expectation. And you never know, you may end up being really talented and reaching that 30-40% annually mark.