>Why they don't [they, i.e. WIND] just report the real numbers?
(My apology to F. Foos. I wrote this before you posted, so please excuse the repetition. What you wrote is correct, but we need to go further.)
Ostensibly, the reason for managing earnings is that analysts might extrapolate from a couple of particularly good quarters and raise future estimates too high. When estimates, or worse, the dreaded whisper numbers, are missed, the stock would get tanked. We have all seen significant companies take a bath after reporting disappointing earnings.
But this isn't the reason.
I believe the real reason for keeping a lid on EPS and taking the mystery out of each upcoming earnings report has to do with managing sales and operations. Understand that investors are not the only stakeholders aware of a company's earnings estimates. Actual and potential customers are equally aware, and they often try to use this information to their advantage, in nefarious ways.
As the end of a quarter draws near, a customer may hold out for a sizable discount or special treatment, believing that the salesperson and the company desperately need the sale to make estimates. In this traditional way of operating, sales tend to bunch up at the end of reporting quarters, reducing earnings visibility by management, analysts, investors or anybody else. This results in the so-called hockey stick - the bane of managers and salespeople alike. If you combine this phenomenon with the occasional optimistic analyst estimates, you have the makings for a company meltdown.
By managing expectations and reported EPS, WIND has been able to eliminate the hockey stick. By making everyone aware of the practice, the craziness has been eliminated. Investors cannot be "disappointed" with earnings known well in advance. Analysts follow management guidance. Customers are never positioned to greenmail the company into submitting to undeserved discounts, in particular the evil "buy out".
>Are they trying to manage expectations or are they constraining real growth?
They very much want to limit expectations to levels that will be obtained without having to sacrifice long-term business opportunities for the company. For example, WIND does not want to earn more than a 30% operating profit, no matter what. They will, but the goal is to continually invest in the business, not surpass MSFT and INTC, or other monopolists, on net profit rates. They do not want to get caught up in the cycle of trying to meet short-term expectations of greedy traders.
They are NOT constraining real growth. They are promoting real growth by investing in the business and eliminating the sales frenzy.
If you are concerned that they tell customers to wait until next quarter, in order to manage earnings by constraining revenues, and thereby turning away business, then rest easy. Customers are not affected one iota. We are only talking about innovative accounting procedures, rather than closing the door on sales. WIND does not postpone actual sales when they delay recognition of revenue.
According to accepted accounting standards, product license income is accrued at the instant the product is shipped. One way to postpone the accrual is to provide the customer an evaluation license rather than a regular product license. The product is exactly the same, but an evaluation license does not count as sales. After the quarter closes, the license is converted to regular form, triggering the sale. The customer is happy to have a short reprieve before receiving an invoice for a product shipped. If the customer happened to pay in advance (which never happens), the money would be classified as deferred revenues.
Another way to level the revenue playing field is to be maximally conservative about allocating royalty and maintenance revenues over a lengthy future. Of course this is how these revenues should be accounted for anyway, but companies hard-pressed to make their numbers often book these revenues as aggressively as allowable under standard accounting procedures, even to the point of bending the rules (witness the recent Informix revelations.)
You might have asked:
>How far can the company go in managing EPS before other problems surface?
Great question. For the first time we saw a bit of a problem this quarter just reported. It is unfortunate that product license revenues had to be constrained to the point that there was an illusion of a slow-down in product sales relative to other parts of the business. The Street can get very nervous, and decidedly pessimistic, if it believes that the lifeblood of a company is threatened or is under newly discovered pressure. To counter this mistaken observation, management had to repeatedly clarify the situation with analysts during the conference call.
One way to handle excess services income, is to issue a warning near the end of the quarter that services revenues are likely to blip up by $X during the quarter due to "bla bla bla", but that this will average out over the following quarter due to backlog adjustments. The analysts would have upped this quarter's estimate by the quantity (0.36 * 0.4 * X / Shares Outstanding) and subtracted it from the following quarter. Had this scheme been implemented during the last quarter, the analysts would have increased estimates by 2 cents, to 18 cents. I would have upped my estimate to 20 cents, which is what the company would have reported.
An even more difficult potential problem managing EPS awaits the ramp up of I2O. I2O will be bumpy, with excessive estimates sometimes, and underestimates at other times. I suspect this is the justification for delaying the inclusion of I2O in analysts' estimates until after a couple quarters of experience with I2O have elapsed. The hope is that future I2O shipments will quickly become predictable, enabling management to provide reasonable guidance.
Unfortunately, it may be virtually impossible for management to provide accurate guidance months in advance, even after experiencing a couple of quarters of I2O shipments. I believe I2O royalties will ramp up much faster than most people expect because the channel for each I2O product will have to be stuffed, generally before there are sales to end users. Once the channels (and there will be all types of them) are full for the initial onrush of products, there will be a lull for a couple of quarters, followed by accelerating growth. Of course this assumes I2O will prove to be as useful as we imagine. In any case, it may be years before management can accurately project I2O shipments.
An alternative approach would be to set up a separate I2O royalty category for reporting purposes, and then, as the quarter unfolds, provide as much information as practical about I2O shipments. Guidance could be provided as usual for the core business, i.e., all but for I2O royalties. No guidance whatsoever would be provided about I2O royalties - other than providing preliminary reports of I2O shipments in "real time" (hard or soft?), along with any guidance provided by Intel and other I2O chip manufacturers. Analysts would be encouraged to enlist the aid of market researchers to project the number of I2O shipments.
Analysts would use the reported I2O royalties to refine the relationship between preliminary shipment figures and the I2O contribution to EPS. Indeed, the I2O contribution to EPS also could be reported separately, if only to emphasize that the core business met or exceeded analysts' estimates.
Now, image what would happen when earnings are reported, if I2O is reported separately as recommended. We know analysts will underestimate the core EPS by 2 cents. If, by chance, core EPS falls short of the estimates, then the stock will be punished, perhaps brutally if concerns surface about the core business. There is nothing new or different about this drill.
If analysts tend to underestimate I2O royalties, the stock will be rewarded. If analysts overestimate I2O royalties, the stock will suffer, but not the company. This error does not lead to a hockey stick effect, because the company is passive about I2O shipments. If analysts overestimate I2O royalties, it is because they made a mistake, not WIND management. Shame on them, not WIND. Perhaps the best solution really is to separate I2O royalty reporting and let analysts earn their money.
Warning to analysts: Estimate I2O shipments accurately or get another job.
Final concern: What happens if I2O flops, and royalties never materialize or, if they do, they rapidly peak and start to decline?
If this happens, it happens. It may as well happen out in the open, so analysts and investors see it, accept it, and get on with all the other positive developments in the embedded systems space. After a few quarters of declining I2O royalties, with no hope of a reversal, I2O royalties would be collapsed into product revenues, as they are reported today.
Allen |
