Hello Reggie,
Thank you for referencing the article "The Problem with Earnings Estimates". I found the article to be right on, and it confirmed many of the views that I have averred here on SI, many times.
As for your reminder to me of the irrationality of analysts, "they are very rational. Remember that they are in place to support sales and banking. Since most analysts aren't lucky enough to be part of the INTC underwriting team, most probably act in support of sales. That means they must generate sales commissions. In order to do that, they must get you to trade, which means they don't want you to sit in a stock and they need volatility to accomplish these means - translating to multiple, and often unnecessary, buy and sell recommendations."
May I remind you, Reggie, that you are "preaching to the choir" here. As an Ex Sell Side analyst, I was well aware of the games that my competitors would play. However, there were some analysts, like myself, that did not ascribe to these "conflict of interest" methods of analysis. In fact, because of my views on the nebulous value of earnings estimates, as outlined in your referenced article, I refused to provide quarterly earnings estimates. Instead I would point out inconsistencies.
For instance, How can Computer Mainframe analysts accurately forecast IBM's earnings for the current quarter without knowing the number of shares that IBM will be buying back this quarter, or the extent of the impact of currency translation on the bottom line, or the exact tax rate that IBM will be using, in Q2.
Another example of the folly of earnings estimates was last week's trading of AOL. When AOL's earnings were announced three weeks ago, it was said that AOL beat analysts estimates by $.02. In the immediate days to follow, AOL's stock went up almost 15%. Then, last week, it was disclosed that included in AOL's earnings were accounting maneuvers that treated a sale of a division as a "lease back sale" rather than a one time gain. This added $.05, or 1/3 of AOL's total earnings, to the quarter. In other words, without the accounting "gimmick" AOL would have fallen $.03 shy of analyst's estimates. Thanks to Jim Krammer, who appeared on CNBC the next morning after the news was released, (he is very long AOL), he put on a "spin" that this news was no big deal. The stock actually closed up 2 3/4 points for the day. "Yeah", he boasted, "I knew about AOL's accounting methods for days before the news came out, So what if AOL was using a less than conservative accounting technique, I would try it too, if I could". But he entirely sidestepped the main point. The analysts had made their earlier earnings estimates without having the benefit of knowing that $.05 from this sale would be used to help boost AOL's quarterly earnings. (in fact, if allowed to hold up, this accounting gimmick will help AOL's Quarterly earnings for the next 20 quarters to come)
But back to IBM, which as you know I have been short for quite some time. You may be interested in reading what I wrote last week.
"For more than a year IBM has been buying back significant amounts of their stock. In most quarters, because of the lower shares outstanding, this has contributed meaningfully for IBM's ability to meet analyst's earnings estimates. Therefore, the year to year quarterly comparisons, going forward, will be very difficult for IBM to maintain without continuing to buy back similar amounts of their stock. With the price of their stock now up more than 70 % from 1997, Q298 will not have as positive of an impact as it did last year. Future quarter's earnings therefore, will be compared to those from the year's before and since both quarters will include the impact of stock buybacks to boost earnings, they will not get the positive effect that they got in the year before.
The price of IBM's stock is up some 50 points from April 97. Yet analysts do not know how many shares IBM will be buying back in any one quarter. How can the analysts possibly hope to provide a meaningful quarterly, or yearly estimate numbers. To make matters worse for IBM, for the past 18 months they have gotten help from a lowering of their tax rate. As I recall, tax rates have come down from 39 % to 31.7 %. Going forward I don't think IBM will be able to get rates much lower. Again this will be bad for quarter vrs. quarter comparisons.
Now I admit Gerstner knows how to put a good "spin" on events, but can he overcome continued currency translation problems, slow growth in Asia, the substitution of Services Division's revenues at 23 % GPM for eroding hardware and software sales at 70 % gross profit margin? And of no small concern, the substitution of Intel's Xeon and eventually Merced chips for IBM's own manufactured hardware revenues, and at the same time absorb the effects from the continued proliferation of Window's NT which is reducing IBM's software revenues?
Regards,
Jules
|
