by Martin Zweig
Buy-and-hold would fare somewhat better if dividends were considered. My estimate for dividends on the ZUPI (there is no precise figure) is about 3.5% a year over the period … somewhat less than dividends on bigger, blue chip stocks, which dominate the major stock averages. When the dividends are added to the capital gains for the buy-and-hold investor, the return increases to 10.2% a year. That would make $10,000 grow to $601,225 over the 42-year test period.
But, we also have to allow dividends of 3.5% during the buy periods of our Monetary Model (actually, the dividend yields would have been a bit better than that because many times our model had us buying near market lows, when yields were greater). That would have increased the total ending value on the Monetary Model to $2,526,419, a nice 14.1% yearly return. In other words, if you had bought the average New York stock on the model’s buy signals, allowed for dividends, and then switched to money market instruments on sell signals, you would have made about 14,1% a year for more than 42 years and would now have about 250 times your original stake.
Transaction costs (commissions) would not have been too important since the portfolio would have turned over only about once every two years. Moreover, you could have traded no-load mutual funds and avoided commissions entirely, although you would have absorbed some management fees. But at least the reinvestment of interest and dividends could have been done with only minimal cost.
I have also ignored taxes. That assumption is warranted in pension accounts, including some you may have yourself such as individual retirement accounts (IRAs) or Keogh plans. Of course, the buy-and-hold investor also has to reckon with taxes sooner or later … whenever he decides to cash in. He also has to pay taxes along the way on dividends.
GRAPH H
Ned Davis Research
The right side of table 10 shows the results of sell signals on the model against the ZUPI. Five times, the broad market went higher on a sell signal (once, just barely); five times, it fell as predicted, including two horrendous losses in the bear markets of 1969-70 and 1972-74. Had a hapless investor stayed in the market during the sell signal periods, he would have lost about three-quarters of his money, showing a loss rate of 6.2% per annum. Had he paid attention to the Monetary Model, he could have avoided such grief.
TABLE 11
MONETARY MODEL VS. STANDARD & POOR’S 500 INDEX: 1954 to 1996
BUY SIGNALS SELL SIGNALS
Date
S&P
% Change
No. of Months
Date
S&P
% Change
No. of Months
3/17/54
26.62
+64.9
18
9/9/55
43.89
-8.0
26
11/15/57
40.37
+42.2
22
9/11/59
57.41
+.6
11
8/23/60
57.75
+54.0
66
3/10/66
88.96
-3.5
11
1/26/67
85.81
+11.7
15
4/19/68
95.85
+3.1
4
8/30/68
98.86
+5.1
4
12/31/68
103.86
-21.1
21
9/21/70
81.91
+31.2
21
6/26/72
107.48
-34.9
29
11/28/74
69.97
+37.4
30
5/31/77
96.12
+13.4
36
5/6/80
109.01
+59.8
41
10/13/83
169.88
-3.2
13
11/21/84
164.52
+54.0
42
5/11/88
253.31
+34.4
18
11/15/89
340.54
+31.5
56
7/12/94
447.95
+35.3
17
12/20/95
605.94
+7.2
3
3/20/96 *
$10,000 becomes: $256,862 318 mo. $9,505
186 mo.
Annualized return = +131% -0.3%
Buy-and-hold return = +7.9%
Table 11 tracks the Monetary Model against the Standard & Poor’s 500 Index. Buy periods alone would have produced annualized profits of 13.1%, while sell periods showed losses of 0.3% a year. By contrast, buy-and-hold gained 7.9% yearly. Again, all ten prior buy signals had gains, with eight of them returning more than 30% each. Five often sell signals led to market declines, and two of the four gains in sell periods were minimal. The Monetary Model is plotted in graph H (pp.76–77) against the S&P 500 Index. The most bullish zone is the area above the upper dotted line. The most bearish zone is the area beneath the lower dotted line.
You don’t have to use my buy and sell signals (6 points and 2 points respectively). You can use the rating on the Monetary Model in conjunction with other indicators to make major market-timing judgments or to make partial moves in the market. For example, rather than using the all-or-none approach of buy and sell signals, you might want to increase stock investments as the model gets better, and to decrease them as the model falls. When the model is neutral at 4 points, you might be 50% invested. If it rises to 5, you might go to 65% invested. A score of 6 might be worth 80% invested, and an 8 might entice you to go 100% long.
In the other direction, a 3 might relate to 40% invested, a 2 to 25%, 1 to 10%, and 0 to 0% invested. These are merely ballpark suggestions, not hard-and-fast rules.
It is illuminating, though, to see how you would have fared only when the Monetary Model reached its extremes of either 0 or 8 points. Table 12 lists all sixteen cases when the model rested at 8, the best possible score. The market, as measured by the ZUPI, went higher fourteen of these sixteen times, an 87.5% success average. Indeed, the two losses were trivial, including a 2.7% loss in late 1975 when the model temporarily fell below 8 points. But three weeks later it rebounded back to 8 and the market skyrocketed. The other loss was a meager .1% during an eight-day stretch in March 1982. Conversely, in nine of the sixteen cases the market gained 16% or more.
In all, had you invested $10,000 only during the 90.5 months when the Monetary Model was at its absolute best rating of 8 points, you would have seen the money grow to $145,471, an an-nualized gain of a whopping 42.6%! A truly conservative person could have then retreated to money market funds the rest of the time, having exposed himself to stock market risk for only 7.5 years of a 42-year period, yet far outperforming the stock market. A more typical investor might opt for the buy and sell signals described earlier, or a “partial” strategy such as that described on page 79.
TABLE 12
MONETARY MODEL RATED +8 POINTS VS. ZWEIG UNWEIGHTED PRICE INDEX: 1954 to 1996
+8 Point Span
Start
End
No. of Months
% Change in ZUPI
6/16/54
to
1/19/55
7.0
+28.4
1/22/58
to
10/17/58
9.0
+32.0
3/17/61
to
6/10/61
3.0
+4.0
4/7/67
to
10/7/67
6.0
+16.1
9/25/68
to
12/2/68
2.5
+7.7
11/3/70
to
7/6/71
8.0
+22.9
11/19/71
to
3/16/72
4.0
+19.7
12/13/74
to
7/28/75
7.5
+48.5
11/5/75
to
12/6/75
1.0
-2.7
12/24/75
to
6/7/76
5.5
+20.2
6/16/80
to
8/26/80
2.5
+12.4
9/21/81
to
2/1/82
4.5
+1.3
3/8/82
to
3/16/82
9.5
-.1
7/26/82
to
5/19/83
9.5
+63.6
12/4/90
to
6/22/92
18.5
+36.9
7/2/92
to
8/18/92 *
1.5
+1.4
$10,000 becomes:
$145,471
90.5 months
Annualized return =
+42.6%
(7.5 years)
Table 13 on page 81 enumerates the ten cases when the Monetary Model rested at its lowest possible score of zero. Stocks fell during seven of those ten spans, for a nice 70% success average. The three times stocks rose, the gains were tiny ones of 0.6%, 2.5%, and 2.85% respectively. The model was in its worst mode for a total of 27.6 months. You would have lost about 40% of your original investment in that span, an annualized rate of loss of 20.3%, fast enough to send you to the poorhouse in a hurry. Obviously, the stock market is no place to be when monetary conditions are hostile. Yet, stocks are super attractive when the Fed is loosening and interest rates are falling. In sum, “Don’t fight the Fed.”
TABLE 13
MONETARY MODEL RATED ZERO POINTS VS. ZWEIG UNWEIGHTED PRICE INDEX: 1954 to 1996
Zero Point Span
Start
End
No. of Months
% Change in ZUPI
10/14/55
to
10/15/55
0
+.6
11/18/55
to
2/5/56
2.5
+2.5
4/17/69
to
6/18/69
2.0
-6.6
5/4/73
to
1/2/74
8.0
-22.0
1/6/78
to
2/29/78
2.0
-.9
6/30/78
to
7/6/78
0
-.9
8/18/78
to
4/16/79
8.0
-8.3
9/18/79
to
12/7/79
2.5
-8.3
2/19/80
to
5/1/80
2.5
-11.4
2/1/95
to
2/15/95 *
0.5
+2.85
$10,000 becomes:
$5,883
27.6 months
Annualized return =
-20.3%
(2.3 years)
CHAPTER 5
Momentum Indicators—“The Trend Is Your Friend”
In the old days all stock transactions were actually printed on ticker tape that rolled out of a machine topped with a glass dome. To this day the activity on the market itself is called tape action. Of course, we now have electronic machines, and the physical ticker tape is almost obsolete. However, every transaction is still reported with the name of the stock, the price of the trade, and the volume.
Any calculation using price and volume is in the realm of technical indicators. Occasionally the sentiment-type indicators, a first cousin to tape action, are included in technical analysis, but they don’t truly fit that category. We’ll cover all the sentiment indicators in another chapter. Right now we’re just talking about price and volume.
Of the two variables, price is more important than volume. For the moment let’s concentrate on price. In the market, as we’ve seen in earlier chapters, you can construct a price index such as the Dow Jones Industrial Average or the S&P 500 or my unweighted index, the ZUPI. You can observe the behavior of a price index and measure tape action from it. For example, you can analyze the change in the Dow Jones average. If the average is up by X%, it’s bullish; if it’s down by X%, it’s bearish. That would be a very simple type of indicator.
To complicate matters, some people might argue that if the market is up by X%, it’s bearish because the market is overbought and likely to decline. Conversely, one may contend that if the market is down by X%, it’s bullish because the market is oversold and ready to rise. That brings us to the key question, Does strength tend to beget strength, or does strength tend to wane and lead to weakness? Let’s look at it scientifically.
After years of testing various market averages, advance/decline ratios, volume figures, and other indicators, I have found that strength does indeed tend to lead to greater strength. Every single bull market that I’ve seen has started with a tremendous rally. The rally doesn’t necessarily come the first day after the bear market ends. Occasionally you have a period of weeks or even months during which the market backs and fills and bounces around in what technicians call a basing action. If conditions are right, a rally eventually ensues.
For a raging bull market, you need falling interest rates, probably an economic recession (that helps the Fed to loosen up and rates to fall), lots of cash on the sidelines, good values in the market—namely, low price/earnings ratios—and a great deal of pessimism because, as we’ll see later, pessimism means there’s an abundance of cash. If all these conditions converge, the market should rally very, very strongly, and the first rally of the bull market should be the best one.
When that rally blasts off, prices should go through the roof. Remember my analogy about launching a rocket to the moon? The rocket must have sufficient thrust to get through the atmosphere and into outer space. The market works similarly. The first rally must have a tremendous surge for a major market advance. If it does, it generates more buying enthusiasm and brings in people who missed the first move. It prevents a large correction because that first rally reverses the market psychology. People who missed it are sitting there loaded with cash and eager to get aboard. So, after the smallest setback, new buyers enter and there is no sharp correction. One of the frustrating things for people who miss the first rally in a bull market is that they wait for the big correction and it never comes. The market just keeps climbing and climbing. It feeds on itself in frenzied fashion and propels prices considerably higher for six months or so, and sometimes longer.
ADVANCE/DECLINE INDICATOR
As our first momentum indicator, let us examine data on advances and declines on the NYSE. Advances comprise the total number of stocks that rise on a given day, and declines those that fall. We’ll ignore the day’s total of unchanged stocks. Here’s how you calculate an advance/decline ratio: If 1000 stocks are up on the day and 500 are down, the ratio would be 2-to-l. Of course, advances predominate when the market does well, and declines predominate when the market does poorly. It is a sign of very strong momentum when advances overwhelm declines for a significant span, and vice versa.
I like to track the Advance/Decline Ratio (A/D ratio) over a ten-day period. It’s very rare for advances to lead declines by a ratio of 2-to-1 over such a span. When that happens, one could rightly say the market’s momentum is strong. The test then would be the market’s performance after such relatively rare events.
Beginning in 1953, there have been only eleven cases when the ten-day A/D ratio reached 2-to-l or
more. There were a couple of other cases, which I consider repeats since they occurred a few months after the first signal. For example, in August 1982 the ten-day ratio surpassed 2-to-1 and did it again two months later. I’m ignoring that second signal because it’s superfluous. The last signal came in February 1991. By press time for this book, the signal had worked very well.
Now let us track the signals, as shown in table 14. The first column lists the dates of these A/D signals. The second and third columns show the percentage changes in the market as measured by the Standard & Poor’s 500 Index and the Zweig Unweighted Price Index, respectively. The first A/D signal came on January 26, 1954. Three months later the S&P 500 was up 7.1% and the ZUPI had risen 5.2%. If you follow the columns down, you’ll see how the other signals performed three months later.
Had you invested $10,000 in the market averages only at such signals, held for three months, and then sold, you would have had $22,068 on the S&P 500 Index and $30,451 on the ZUPI. Bear in mind that that’s in a cumulative period of only two and three-quarter years, eleven separate three-month periods. The rate of return per quarter was 7.5% on the S&P and 10.6% on the ZUPI. The annualized returns would be more than four times the quarterly figures when compounded.
The right-hand columns in table 14 show the market’s performance six months after these signals. Note that the S&P 500 and the ZUPI were up at least 10% six months later in virtually every case. You would have more than quadrupled your money in the S&P and sextupled it on the ZUPI. The return per six months is 15.2% on the S&P and 19.2% on the ZUPI. Those returns are extraordinarily high in the stock market.
