re: I find some intelligent going forward
view of price/sales is probably more helpful.
Once I realized all the games companies were playing with proforma results, I started to rely more of P/CF and P/S. However, with all the vendor financing and bad loans in so many industries, even historical sales numbers are suspect.
How can an investor use P/S to find Value (= set a buy-price at a level where the P/S is likely to go up in the future)? One way is to find the historical range the company has usually been in. For instance, AMAT usually bottoms at a P/S of 1 or 2, and peaks in its cycle, at a P/S of 4-5. Buying when the stock is at a P/S of 1-2, has always been an excellent investing decision.
However, there are several problems with this method. First, you have to use a long time-line. Valuations (for P/S, and every other metric) got stretched, beginning in 1995, and got into levels never seen before (not even in 1929), in the late-1998 to early-2000 time period. If you use that recent data to set the range, you are using data when conditions were in a twice-a-century mania. That data should be thrown out, as an "outlying data point, not representative of the longterm trend". For instance, I doubt AMAT will ever (at least in my lifetime) hit a P/S over 10, as it did in 2000.
And, if you use earlier data, you get into trouble, because the industries and companies have changed so much. For instance, CSCO was the biggest company in network equip in 1995, while now it is an upstart in the (far bigger) telco equip industry. And the internet has changed CSCO's prospects more than a little, from 1995 to now. So, using P/S data from 1995, for CSCO (or JDSU, whose industry has changed just as much) is an apples-to-oranges comparison.
Then, there is the problem of using historical P/S ratios, in companies that have grown and changed dramatically through acquisitions. I remember, last year, looking at the P/S range for WCOM, for each of the last 10 years. The numbers were all over the map. Then, I realized that WCOM's many acquisitions meant it was a very different company each year. Another apples-to-oranges comparison.
OK, if I can't use P/S in a relative way (using the historical range to decide over/under-valuation), can I use it in an absolute way? Well, doing that means I have to know what margins are going to be, going forward. Sales, and any other number measuring a company's performance, are only important to the extent they affect profits. It all comes down to profits (real profits, not pro forma "profits"). The trend in sales is a proxy for the trend in profits (and hopefully a more reliable number, as the Creative Accountants have a harder time playing with the sales numbers). But this is only true if the trend in margins is known. Increasing sales was used as a reason to bid up AMZN during the mania, and this worked as long as everyone ignored the fact that AMZN lost money on every book they sold.
So.......JDSU's margins will be what, going forward? I don't have a clue, and I don't even know any way to try to get an answer. Do you? |
