$XVG and the Four Percent Model

In Martin Zweig's book Winning on Wall Street, he wrote about his friend Ned Davis' Four Percent Model. Davis used the Value Line Composite Index (symbol $XVG on Stockcharts). His model was simple:

1. Buy if the $XVG weekly close rallies 4%+ off the most recent weekly swing low close.

2. Sell if the $XVG weekly close corrects 4%+ from the most recent weekly swing high close.

Note: you can also use the sell signals to short, and the next buy signal to cover.

This model is designed to force you to stay with the market trend. When the market is choppy and not trending, it's prone to whipsaw.

Using this model, we got a buy signal on the February 19 weekly close at 312.80. The most recent $XVG swing high weekly close was 361.46 on Friday, April 23. We'll get a sell signal on a $XVG weekly close under 347.00. Today, $XVG closed at 340.37. We may get the sell signal Friday afternoon.

Zweig published historical results of the model from 1966 to 1995 assuming 100% invested at all times, switching from long to short. Short trades produced winners at 46% with payoff odds of 2.7:1 (9.5% winners to 3.5% losers). Long trades won 49% of the time with payoff odds of 3.6:1 (14.1% winners to 3.9% losers).

Looking at Zweig's charts, if we had just taken the short signals during the stock bear market from 1966 to 1982:

21 winners
17 losers
55%

9% avg. win
3.9% avg. loss
2.3:1 payoff odds

If we had just taken the long signals during the stock bull market from 1982 to 1995:

12 winners
12 losers
50%

15.5% avg. win
3.3% avg. loss
4.7:1 payoff odds

So, optimizing the strategy depending on the nature of the secular market produced a significantly better won-loss percentage on short trades at a cost of reduced payoff odds. Long trades saw the same win percentage, but much higher payoff odds when confined to bull markets.

The problem is you can't buy the Value Line Composite Index. Fortunately, SPY is very closely correlated with $XVG.