The forthcoming IPO of Pure Storage is highly anticipated by those of us who follow the storage industry closely. The valuations assigned to companies prior to IPO are based on what a very few people, its early stage investors, believe the company is worth. They tend to be pretty optimistic people, because you don’t invest in high-risk things like startups without being a bit of an optimist about getting your money back, so how realistic are those valuations, really? An IPO provides an opportunity to find out.
Prices are tricky things, because they’re an artificial construct; there is no naturally occurring, objective way to know what price anything should be. There are myriad ways our impression of the ‘correct’ price for something can be influenced in ways that are far from rational. The excellent book Priceless: The Myth of Fair Value by William Poundstone, contains dozens of ways our impression of what a price should be can be manipulated.
So how do you put a price on a company?
I spoke to valuation expert Drew Dorweiler, a Board Member of The Appraisal Foundation, who explained that aside from looking at the fundamentals of the business (such as information contained in the S-1), we can also attempt to calibrate our expectations from what other, similar, transactions looked like. Then we make various adjustments to those similar listings, trying to account for the differences between Pure and other companies. The more similar these transactions are to Pure’s IPO, the fewer adjustments we need to make, and the less uncertainty we can have that our estimates are somewhere close to what might actually happen.
“We would want to compare Pure to five to seven other transactions, at least,” he said. “We would look at the multiples implies by these transactions, and adjust based on the market specifics relevant to Pure.”
One pretty good comparison point is probably XtremIO, purchased by [entity display=”EMC” type=”organization” subtype=”company” active=”true” key=”emc” ticker=”EMC” exchange=”NYSE” natural_id=”fred/company/1441″]EMC[/entity] in 2012 for $430 million in cash. Back then, XtremIO didn’t have a generally available product, but Pure has been selling flash arrays for several years now. The architectures of the products are quite different: Pure is an HA-pair architecture, while XtremIO is a full-mesh cluster architecture. How do you adjust for these differences?
It requires deep knowledge of the market, and an ability to analyse what makes a reasonable adjustment assumption and what doesn’t. We also need to account for what the future might hold, because the present value of a company should, in theory, be equivalent to the discounted value of all future cash flows.