Valleywag
Technologists are instinctively averse to debt. The cycles are too swift and mistakes too punishing, the conventional wisdom says, to subject a startup to the burden of debt; cash is better spent on growth opportunities than interest. But Facebook has never followed the usual script for a startup, and its CFO, Gideon Yu, is no herd-follower, either. No wonder that the news that Facebook is leasing $100 million worth of servers, after raising a $360 million round of venture capital from Microsoft and Li Ka-Shing, is causing such a ruckus — and some misconceptions. Here are the instant myths that have arisen:
- Facebook is borrowing money. Actually, no. Facebook's $100 million is a lease, not a loan. Facebook is on the hook for rent payments, and will have the option to buy the equipment at the end of the lease. But technically, it's not on the hook for the $100 million. That said, it's hard to imagine how, at the end of the lease, Facebook functions without the servers it's renting. What Facebook is doing with the lease is delaying the day it has to buy its servers outright.
- Facebook is borrowing money because it can't sell its equity. Henry Blodget of Silicon Alley Insider put forward this theory. In moments like these, one wonders less why the former stock analyst was barred from the securities industry. A venture lease like Facebook's usually requires an "equity kicker" — warrants to purchase stock, the corporate version of the stock options employees get. In this case, Facebook negotiated a warrant-free deal, but it's given TriplePoint warrants in the past. So Facebook's lender, TriplePoint, is betting that Facebook's equity is worth something; it wouldn't have offered it leases in the past otherwise. A warrant-free deal may not be great for TriplePoint, but it's great for Facebook.
- Facebook won't know what to do with all those servers. This is the most nonsensical theory of all. Facebook isn't renting servers for the workaday business of displaying webpages and photos. While it needs some more servers to keep up with site growth, the main reason it needs so many servers, I believe, is that it's planning to use them on the massive job of analyzing data to target advertising and content to its users.
- Gideon Yu must be an idiot. We've ribbed Yu for being a bit slick. But as YouTube's CFO, Yu negotiated the company's sale to Google for $1.65 billion. Shall we not give him some small amount of credit for knowing what he's doing?
And as for that bromide about tech companies not incurring debt? In the form of leases, they do it all the time, from landlords and hardware makers. Even startups lease their offices, and IBM, HP, Dell, Sun and others have large financing arms that help put equipment in customers' hands without money upfront. Save for the size of this deal and the name of the renter, TriplePoint's lease would have been wholly unremarkable.
There is one question that remains: How TriplePoint will come up with the money. This is reportedly the largest deal the lender has ever financed. Let's take TriplePoint at its word, though. The company's website says it has says it has "the capacity to put more than $1.5 billion to work" — curiously hedged wording. Note that TriplePoint doesn't actually say it has $1.5 billion on hand to lend. But that's TriplePoint's problem, not Facebook's.
Comments
"...The main reason it needs so many servers, I believe, is that it's planning to use them on the massive job of analyzing data to target advertising and content to its users."
Wouldn't it be smarter to start out small; maybe experiment with 100,000 users and not the entire 100M?
If their revenue was cranking the way great companies tend to they wouldn't need $500M over a 12-month period. Their revenue is weak and their expenses are insanely high IMHO. It appears that they have issues. Remember back a few months ago when young Mark Z. did a public conference call to announce that, among other things, they were going to have a disappointing 2008 in terms of revenue? Who does that? Why? I think it's because the revenue isn't there.
Yeah, this spin doesn't ring true.
It's not like they're doing so well in terms of revenue that they have to feed the beast. We all know that Beacon was a failure. In my mind that is the real story here.
@sample032 : some problems dont lend themselves to partitioning. 10,000 random people may have no friends in common.
More likely, this is a bullish sign. They've probably done the offline targeting analysis and proven that it works and now need the servers to be able to do the same in real-time. Or they are having more intl. user-growth than expected.
@xyahoo: I never said random. Start with 20 random users and do a breadth-first traversal of the friends graph until you have 10,000 users.
Facebook.com sits on a goldmine in userdata that they are not capitalizing.
What is wrong with social networks todaY? How in the world can you not make money from BILLIONS of hits?
Seriously!
A smart piece of financial journalism. Blodget was out to lunch on this one.
"Shall we not give him some small amount of credit for knowing what he's doing?"
You must be new here.
Facebook had a great "surprise" business model. Problem is the surprise is over and everyone is busy copying it before Facebook has even monetized the thing. It would be as if Google had announced three years in advance that they were going to revolutionize search and give away gigs of storage for e-mail, photos, videos, etc. and then only implemented enough of the concept to show that it could be done. Google let its actions speak louder than its words.
"Social" is well understood now by anyone who had doubts. The whole thing is in middle age at this point. Myspace has caught up, even slow as molasses (on changes) Orkut has caught up, and most of the things I do with Google now have social aspects added to them.
It is ridiculous, having totally "telegraphed their punch" that Facebook is still building out its infrastructure.
It's over for this outfit.
Not to defend Blodget here, but perhaps Facebook wasn't willing to price those warrants at what a non-rose-colored-glasses finance company would find as a fair value.
Isn't Microsoft already on record saying they overpaid? If so, what would that do to all the BS money everyone thinks they have due to these private placements?
PS: You're right about a lease not being a loan. Technically.
But I've worked at start-ups that leased equipment. You can end up in an interesting catch-22 where you need more business to make payroll, but you need more computer power to support that business (i.e. you underestimated), but your lease payments for what you have are huge already... back to the leasing company to beg for a new deal. Guess what... these guys probably know the status of your business better than you do. This is how IBM operated for many years, and still does in fact, even though they have a lot of competition now thanks to the PC phenomenon.
"Facebook isn't renting servers for the workaday business of displaying webpages and photos."
Oh yeah? That's pretty much what they've done with all the other servers they've bought in recent history.
You're all wrong... This is a tax play for Facebook.
Leasing vs. Purchasing
Lease = deduct all payments made immediately thus offsetting income.
Purhcase = capitalize over the life of the asset. This means that instead of being able to deduct the entire amount up front, they instead must take the $100 Million and divide by the average number of years it'll be placed into service.
Also, I'm sure these servers will be all but obsolete in 5 yrs so they won't exercise the option to purchase anyway. Better to start over at the end of the lease term with the latest equipment and expense all those millions of dollars in payments once again.
Not every decision FB makes has to do with some grand strategic play...it's a simple matter of taxation and time value of money.
@sample032: and what makes you think they haven't already done that? And even done it with a larger group? And it worked?
Has anyone tried to figure out how to make money with social sites?
Start a discussion:
Login with your username and password below. Or comment on this post via email.
Forgot your username or password? New User?