Oscillator Tuning

02/12/01 04:22:20 PM

by Dennis D. Peterson

Your selection of indicators needs to fit a stock's personality.

Security: MSFT
Position: N/A

Each stock has its own personality in terms of price and volume behavior and goes through different phases of accumulation, consolidation, price volatility, trends, and is sensitive to time periods. Using the same indicator all the time is akin to saying, "I think the stock behaves in the same way all the time."

In addition, many indicators are similar. In his book "The New Technical Trader," Tushar Chande points out the following statistical correlation between indicators (correlation - indicator/ pair): 93% momentum/RSI; 78% momentum/Stochastics; 93% momentum/CMO; 77% RSI/Stochastics; 82% CMO/ADX.

What would happen if you made adjustments along the way as behavior changed? If a stock is in a trend, then trend following schemes such as moving averages work best. Any scheme employing moving averages will be inherently slow to react because moving averages lag. However, the lag from moving averages is mitigated by the fact that trending stocks tend to stay in a trend. What happens when the price starts to rise and fall quickly? Volatility is the nemesis of all traders unless you use stops --- and even then there are losses.

Microsoft (Nasdaq:MSFT) makes a good case study. For years MSFT had a slow steady gain. From 1990 though mid 1998 there was a slow upward climb. If you use a simple moving average crossover trading system (ignoring your gain due to stock splits), optimize on an exponential moving average period, use daily data, a $1,000 investment in 1990 rises to $1,938 by 1/14/92, but then falls to $772 on 3/28/94. However, if you use weekly data and optimize, you have about $2,200 on 3/28/94 by investing $1,000. This shows that price period makes a difference.

The weekly chart of MSFT shows that from 1990 to 1992, prices gradually rose. There is a period of accumulation. Edwards and Magee's "Technical Analysis of Stock Trends" states that accumulation is "the first phase of a Bull Market. It is the period when farsighted investors begin to buy shares from discouraged or distressed sellers." Accumulation is a period of relative price stability as the stock is being quietly purchased (Figure 1: second chart from bottom). For this reason I prefer negative volume index (NVI) as a trading signal, since it indicates what the smart money is doing, that is, buying on the days when volume is down. Using an NVI crossover trading system, a $1000 investment in 1990 is worth $2,890 on 3/28/94 using daily data.

This NVI crossover trading system buys/sells when the NVI crosses above/below its own moving average (of opt1 periods: of optimized periods). The Metastock trading formulas are shown below. For the benefit of those who haven't used them before, enter a long position when NVI is greater than the moving average (Mov) of NVI, (based on close) using an optimized number of periods, opt1, and an exponential moving average.

Long entry
NVI() > Mov(NVI(), opt1, EXPONENTIAL)

Close long entry
NVI() < Mov(NVI(), opt1, EXPONENTIAL)

Short entry
NVI() < Mov(NVI(), opt1, EXPONENTIAL)

Close short entry
NVI() > Mov(NVI(), opt1, EXPONENTIAL)

Figure 1: Microsoft volume and price (bottom two charts), with $1,000 equity growth using an NVI crossover trading system (second chart from top) and stochastic RSI trading system (top chart).

Graphic provided by: MetaStock.

Graphic provided by: Data vendor: eSignal<.

What period should you use for NVI? Between 1990 and 1992 the best choice for equity performance was 200 days, but coming in close were periods of 160, 180, and 170 days. For the entire 11 years the best results were 170 and 180 days. The implication is that if you optimize every nine months or so, your answer is going to be in the range of 160 to 200. I find 170 coming up often in today's market. The equity curve in Figure 1 is based on an NVI using 170 days. While you could optimize every nine months, and achieve better NVI results, the method is still valid (using a 200-day NVI period from 1990 through 1997 results in making $600 less than if you had used a 170 day NVI).

In 1997 you can see some volatility, and in the beginning of 1998 a jump up in prices and then a fall. A telltale sign is the difference between the top and bottom Bollinger bands using weekly data. Up to 1997 there was very little difference and then starting in late 1996 and going forward the differences start to increase. The implication is that Microsoft is changing its behavior and seeing changes in momentum. RSI is a possible candidate for a new indicator. However, a better choice is StochRSI (see posting on 12/11/2000).

If you were to start in 1998, with a simple trading system, that only considers long positions, based on StochRSI being greater than a value, and exiting when StochRSI is less than a value, the result is that the $10,000 you had at the beginning of 1998 now increases to about $35,000. In contrast you would have $16,000 by staying with NVI using optimized parameters from 1998 to today's date.

If you select moving averages you are better off using weekly data to fit the characteristic lagging behavior. NVI works well with accumulation phases, when the trend is not strong. When volatility increases, it is time to consider changing indicators. RSI, as opposed to Stochastics which is designed to support a price channel, works well in a phase characterized by momentum swings. StochRSI is even better than RSI as a pure momentum indicator.

Dennis D. Peterson

Market index trading on a daily basis.

Title:	Staff Writer
Company:	Technical Analysis, Inc.
Address:	4757 California Ave SW
	Seattle, WA 98116-4499
Phone # for sales:	206 938 0570
Fax:	206 938 1307
Website:	www.traders.com
E-mail address:	dpeterson@traders.com
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