|In earlier postings I have often mentioned using 52-week new highs and new lows as an indicator to see if a bottom has been reached. In a previous posting I used a ratio with thresholds to determine bear or bull sentiment. In the current bear market, we are looking for the ratio to move from below a threshold to above it. Today's action, 12/28/2000, saw some improvement in the ratio of Nasdaq new highs/(new highs + new lows) with the number of new highs reaching 160 and the new lows at 279. On 12/20/2000 the number of new highs was 53 and the number of new lows 926! |
To illustrate the validity of using thresholds and looking for an expansion of new highs to new lows, I have created a simple pair of entry and exit criteria for going long. They use only the number of new highs and new lows. Expansion was also used to exploit the new high/(new high + new low) ratio indicator.
The strategy is to enter if the moving average of new highs exceeds a predetermined amount and if a current moving average is greater than an older moving average by a predetermined amount. Use the same strategy for exiting, except substitute new lows for new highs.
The entry rules are as follows (the formulas are in Metastock code and opt1, opt2, etc. are variables to be optimized)
the first condition, cond1, requires the number of new highs, nh, to be greater than an exponential moving average, using opt1 periods, plus some additional amount, opt2,
The second condition, cond2, requires this same exponential moving average to be greater than the exponential moving average starting an optimum number of days, -opt4, ago by some margin, opt3. Again, the second condition compares two moving averages separated in time by a number of days ago (i.e. -opt4), and I want the moving averages to differ by some amount (i.e. opt3).
Finally, I want to enter a long position if both the first condition, cond1, and the second, cond2, are true.
cond1 AND cond2.
In short, the entry condition is for the new highs to start to expand. Moving averages delay seeing the event, but at the same time remove noise from the measurements. The two variables, opt2 and opt3, act as thresholds. If opt2 is close to zero it means enter as soon as the current number of new highs, nh, is greater than the moving average. If opt3 is zero it means enter as soon as the most recent moving average is greater than an average based on older data and then enter.
The exit strategy is the mirror image of the entry strategy, simply substituting new lows, nl, for new highs, nh.
cond1 AND cond2.
In short, the exit condition is for the number of new lows to expand above a moving average and at the same time for the difference in two moving averages to exceed a certain amount.
|Figure 1: Nasdaq versus $1,000 investments using historically optimized parameters (middle chart) and parameters optimized over the last 14 months (bottom chart).|
|Graphic provided by: MetaStock.|
|Graphic provided by: Data Vendor:DBC.|
|I ran two partial optimizations (full optimizations would increase profit) because the compute time is rather large to do more than 60,000 combinations. The first optimization was for the last 11 years of the Nasdaq, and applying those values of opt1 through opt8 to just the period from 10/23/99 to the present, resulted in a $1,000 investment growing to about $1,300 (Figure 1: middle chart). However, optimizing for just the last 14 months and using those values, the investment grew to $1,500 (Figure 1: bottom chart). In comparison, if I had invested $1,000 in a buy and hold strategy from 10/23/99 to the present, I would now have $890, a difference of 68% between a buy and hold strategy and the strategy of using new highs and new lows. Despite what you might hear to the contrary, capital preservation is not poor investing. It's tough to invest when you have no capital. |
Examining the two equity lines reveals a number of interesting points. First, in a strong uptrend such as the Nasdaq saw from October 1999 through March 2000, using new highs and lows kept you in the market for long positions. Second, it also got you out of the market fairly quickly during downtrends, and the values based on optimization over the last 14 months returned more profit than values based on the historical optimization. When you look at the number of periods for the moving averages, opt1, it went down from seven days, for the 11-year time span, to three days for the 14-month time span. Why?
The volatility of today's market is greater than its historical average, which means that to make more profit you must be quicker to react. The same thing happened to the additional amount added to the moving average in condition 1, namely opt2, as it went from 60 down to 15. For condition 2 the days to look at an older moving average, opt4, went from four days back for the 11 year and down to two days back for the 14 month time spans. For the current market the best values appear to be, for opt1 through opt8, 3,15,6,2,80,30,5. But frankly, you are looking at a relatively small difference, i.e. 14%, between the investment gain from historically optimized values and those optimized for the current 14 months. The significant differences between historical values and the current market are in opt1, opt2, and opt4, which have been noted above.
Using these formulas protects long positions. The second formula is the difference between two moving averages, making it a momentum indicator. Thus, entry for longs is based on the new highs having a positive momentum. Exit for longs is based on the new lows gaining in momentum. Market breadth is a valid indicator for gaining insight into market momentum in terms of favoring long or short positions. If I look at the Dow Jones I see new highs often exceeding new lows currently, and not surprisingly I see the DJIA (Dow Jones Industrial Average) holding ground and gaining a bit. If going short is not something you feel comfortable with, then the Nasdaq in its current mood will be difficult to deal with. Especially with the bulls ready to spring a bull trap (see posting on Pivot Points) when the opportunity arises.
Using the ratio posted on 10/13/2000, if the upper threshold at .7 is reduced to .6 then $1,000 invested over the last 14 months using the same 10-day simple moving average and looking at a five day difference in ratio values results in $1,110. Changing to a 12-day moving average, using an 11-day difference in ratio values, 0.4 for a lower threshold and 0.6 for the upper results in $1,254. In other words, the ratio can give reasonable results with a minimum of computation. The market requires that you adjust thresholds. Optimization is a mechanical way of achieving the same result instead of reading the chart. Your judgment in reading a chart is much more valuable than a tool. Of course further adjustments to the optimized values will be needed to get the last dollar out for tomorrow. But the validity of the approach is still intact. When you look for the ratio to move above a threshold you are looking at momentum. You want ratio momentum to start to increase to use long positions. The thresholds simply provide a level of assurance.
|Company:||Technical Analysis, Inc.|
|Address:||4757 California Ave SW|
|Seattle, WA 98116-4499|
|Phone # for sales:||206 938 0570|
|Fax:||206 938 1307|
Traders' Resource Links
|Charting the Stock Market: The Wyckoff Method -- Books|
|Working-Money.com -- Online Trading Services|
|Traders.com Advantage -- Online Trading Services|
|Technical Analysis of Stocks & Commodities -- Publications and Newsletters|
|Working Money, at Working-Money.com -- Publications and Newsletters|
|Traders.com Advantage -- Publications and Newsletters|
|Professional Traders Starter Kit -- Software|
Click here for more information about our publications!