Donald - I was struck by the apparent complete lack of correlation between the % change in stock price since 9/21 and the earnings estimates change for the same period.
I have one possible explanation for this discrepancy...though I'm aware that an easy 'explanation' for such weird data may never surface. My explanation is based on the criteria used to describe a 'year' in the earnings world. A problem arises, particularly at this time of year, with 'fiscal year' definitions for each company. Consider the following sampling from your group:
'This year' for ASML, COHU, HELX, MTSN, NVLS, TER, VECO ends Dec., 2001 while 'next year' ends Dec., 2002
'This year' for DPMI, KLAC, LRCX ends Jun., 2002 while 'next year' ends Jun., 2003
'This year' for KLIC, PHTN, PRIA, SMTL ends Sept., 2002 while 'next year' ends Sept., 2003
'This year' for AMAT, CMOS ends Oct., 2002 while 'next year' ends Oct., 2003
[there is also at least equipment company whose FY starts in March and at least one in May]
Besides the obvious difficulty in comparing earnings estimates with such an array of fiscal years, there is a particularly thorny problem with the first group (fiscal year December through November). The estimates for these companies will appear to be worse than companies with 'earlier' FYs: stocks in the December FY group currently have 'this year' described as virtually synonymous with CY2001 and 'next year' as synonymous with CY2002. OTOH, companies with an earlier FY have 'this year' described as occurring mostly in CY2002 and 'next year' as mostly occurring in CY2003. And since a recovery is expected to have its most significant impact on earnings in 2003, those stocks whose 'next year' is ~2003 are at a significant and illusory advantage to those whose 'next year' is 2002.
Thus the fact that KLIC expects to earn 0.73 'next year' (10/02 -> 9/03) while NVLS, during a supposedly similar 'next year' (but is actually 12/01 -> 11/02) is expected to lose 0.15 is obviously problematic. Clearly this is an inappropriate comparison. In this example, NVLS' lack of earnings is very likely to be a gross underestimation of actual earnings to occur during the period equivalent to KLIC's fiscal 'next' year.
In my view, unless there's a way to standardize the year-to year comparisons, especially in an industry as cyclical as the equipment sector, the earnings estimates comparisons reside on terra NON firma.
Non of this detracts from my oft expressed and ongoing appreciation for all the work you do for this thread and for SI. |
