Follow the Money!
July 18, 1999
Follow the Money and you will get to the truth.
In last week's post entitled "The Demographics of Stock Investment" , I noted that the ratio of retirees to investing age workers stops falling in 1999 or 2000 and begins rising thereafter. At a minimum, we can expect investable down moves beginning no later that September 2003, and probably earlier.
I noted that the wealth effects of the stock market might prompt retirements at earlier ages than the demographics of age groups might suggest, perhaps accellerating the shift from baby boomer demand to baby boomer supply. I noted some evidence of that shift in the article, namely, the narrowing of the market, demonstrated by the poor performance of the advance decline line for the past 12 months, and a divergence in Joe Granville's on balance volume indicator for the past 6 months.
In addition, that earlier post set forth data on mutual fund stock purchases in two seprate tables that may appear, on first glance, to conflict. In the table below, I set those data side by side for comparison. The second column sets forth Federal Reserve data on equity purchases by mutual funds themselves, while the third sets forth the equity mutual fund inflows or purchases for each year published by the Investment Company Institute.
Fed. Reserve Z-1
The federal Reserve data show that actual purchases of stock by mutual funds have been declining for the past two years from their peak in 1996. In contrast, mutual fund purchases have declined for only one year from their peak in 1997. One explantion for the difference between the figures is management fees and sales expenses. The Investment company institute values stock mutual funds at $3.2 trillions. Assuming that the average expense load is 1% then the first $32 billion of inflows each year goes toward fees and expenses. Thus, the higher the stock market goes, the larger the fixed management fee burden becomes. Inflows must grow to sustain management fees. Throw in another $50 billions of fees for management of the $5 trillion or so of stocks held by pension funds, bank trusts, and "financial planners" and you are talking about a net of $80 billions to $100 billions that must flow in each year just to meet management fees based on current prices.
Wall street doesn't like to talk about this, but very small flows at the margin can move prices a great deal.
A much more timely piece of evidence arrived this week. We had new all-time highs in the S&P 500 and the NASDAQ on July 16, but the 8 day moving average of volume crested a week earlier at a much lower level than during the April highs and the January highs as well.
In addition, the 8 day moving average of NYSE and Nasdaq volume was lower on June 15 than at the previous low on February 19, 1999.
In a healthy bull market, new highs in price are accompanied by new highs in volume. Corrections that occur on low volume should produce a series of higher lows in volume, just as in price.
New highs in price accompanied by lower highs in volume are a sign of imminent trend reversal. It is an irrefutable sign of flagging demand for stocks.
Here is a chart of the NYSE volume peaks and troughs that correspond to highs and lows in the market:
Here is a similar table for the NASDAQ:
In addition, three panic corrections produced peaks of volume higher than the previous peaks associated with rising prices. These occurred in July of 1996, October of 1997 and August of 1998 and are not listed in the above table. However, these climactic volume peaks merely confirmed the increasing bull market demand for "buying the dips".
Recently, Maria Bartiromo, a spokes-model for CNBC business news asked rhetorically why anyone should care what interest rates are as long as you can get 20% a year on stocks.
The short answer is in the charts above. Modest rises in interest rates mean that capital is migrating from Wall Street to Main Street. Falling demand makes it impossible to continue earning 20% a year on stocks.
I mentioned in the July 6th post that measures such as record high PE ratios, record low dividend yields, and record high price-to-book ratios are not predictive. Indeed, these measures are nearly certain to move higher as long as volume is increasing - that is, as long as we get higher peaks in volume with each new high in market prices.
So then the question becomes, when will this change in volume trend begin to affect the price trend?
To answer that question I decided to follow the money with a new study.
The Federal Reserve data set forth in my post of July 6, 1999 demonstrates that corporate stock buy-backs play a dominant role in demand for stocks. I also noted that the high tech sector companies write puts and buy calls based on inside information to finance these buy-backs.
Well folks, when do corporations usually have reliable inside information? When is it safe for them to write naked puts and buy calls?
The answer is obviously in the last month of the quarter, when they have a good idea of what the quarterly results will be. If the non-public information looks reasonably good, the corporations will write their puts and buy their calls first. Then they start buying back their stock, driving its price higher and forcing the call writers (who are short the stock) to cover, thus driving the price up further.
Corporate executives have a short window in which to exercise options following the earnings release. This means that price ramps which begin in the final month of the quarter will run through expiration week of the month following the quarter, during the height of earnings announcement season. Thus, the executives benefit directly from the timing of this activity. Further, much of the option activity at the beginning of each ramp can be hidden in the fury of triple witching that occurs in the final month of each calendar quarter.
Given the obvious incentives at work, I decided to do a study. Using the 5 day stochastic oscillator I would buy the first oversold alert occuring any time between two trading days prior to to final month of the calendar quarter and extending to the triple witch expiration as an entry point. I would then exit at the first break in the 5 day stochastic oscillator overbought reading occurring anytime within the 8 trading days prior to option expiration of the next month.
My what a surprise!
Here is a table with results for the S&P 500:
One hundred dollars invested during the ramps compounded to $215. One hundred dollars short between ramps compounded to $109.
Below is a similar table for the NASDAQ 100 Index:
One hundred dollars invested during the ramps compounded to $312. One hundred dollars short between ramps compounded to $118.
Given the trend in mutual fund inflows, as well as the half trillion dollar per year supply of stock from "households" at these prices, I would expect corporate speculation in options on their own stock, and stock buy-backs to continue to dominate the investment landscape until the demographic reality finally sinks in. That moment of recognition is years away. Thus, I expect the quarterly pattern to remain in force.
Between July 19 and August 31, I will be looking for a mild downmove in both the S&P500 and the NDX with several 3-5 day trading opportunities in both directions. It will be a "data sensitive" period, leading off with Greenspan's speech on July 22, and ending with the PPI and CPI numbers for July to be announced in August, followed by the August Federal Reserve meeting. Once earnings are out of the way, the market will be on Fed watch. Most important for our trend recognition will be whether volume makes a lower low towards the end of August. If it does, then we have even more reason to expect a Bear market.
Stress is building in the financial system in Latin America. Currency values are a problem in China. Thus, an unexpected calamity could turn an ordinary gentle downdraft into a crash. But I wouldn't bet against the end-of-quarter ramp in hopes of an unexpected catastrophy.
The critical event will be the next quarterly ramp beginning in September. A failed ramp with a lower high in volume would be a sign that the bear market has arrived. The September 1999 ramp will depend on PC sales. Many large companies have announced that they will freeze spending on computers during the last 6 months of the year so that no new changeovers will draw resources away from Y2k monitoring.
The Y2k problem has been an enormous bonanza for the tech industry over the past 2 years. So far in 1999, it looks like the bonanza continues as the small businesses are getting scared that their older pentiums are not Y2k compliant (every pentium motherboard built in the last 5 years is Y2k compliant) and are buying them like there is no tomorrow. The marketing arms at Dell, HP and others seem to be able to find many smaller businesses with weak internal IS departments who they can scare into buying more machines at premium prices. In addition, individuals keep snapping up the latest high-end machines (with expensive CPUs hopelessly IO bound to slower components in the system) as if they were status-laden foreign sports cars, with little thought to the economics of the purchase.
I will be short in August, waiting to pull the trigger and switch to long in early September. The longer we go with no pre-announcement of bad news from Intel, the more I will feel pressured to switch to long. If we get a serious tech pre-announcement, a failed ramp will follow which I expect to take 30% off the NDX.
The failure of the volume trend casts doubt upon the ability of corporate CEOs to continue having their way with this market. Unless the volume trend changes within the next three months, a bear market is imminent.
Occasionally, my correspondents ask me why I am willing to post this sort of information in public. The reason is simple. The bullish psychology is so deeply ingrained that very few will believe the facts. Thus, publication is harmless.
And it fulfills the first requirement of Sir Arthur Keith's dual code, the sharing of valuable information within the group.
It is a start.
Back to Money and Markets Page
(c) 1999 Yggdrasil. All rights reserved. Distribute Freely.