This page has been archived and commenting is disabled.
How Goldman Sachs Gets Stocks It Underwrites on Its “Conviction Buy” List
This is from Stone Street Advisors
I've been analyzing YouKu.com (YOKU) for the past three weeks, and as I said most recently, I can't possibly see how the stock could be worth more than about $21/share, using what I think are pretty optimistic assumptions. I read Goldman Sachs' - YOKU's lead cheerleader
underwriter - report before I published, and I'm still going through it
and finding data, figures, conclusions and other "analysis" I find very
puzzling, to put it nicely. I'd made a note to myself during my first
read-through to go back and check GS's WACC
calculation, since their Cost of Capital was only 12%, which seemed
pretty low to me considering we're trying to value a high-growth,
relatively high-risk company in China.
Today, I finally went back
and checked GS' work, and what I found is at the very least EXTREMELY
confusing, and at worst, deeply, deeply troubling.
Here's why:
GS is using 6.5% as the Equity Risk Premium
for a high-growth, high-risk company with all of its operations an
emerging market economy (at least recently) plagued by poor internal
controls/accounting/audits, shoddy corporate governance, and fraud? In what
alternate reality does that make sense?
Why would we use the U.S.
ERP (6.5%, which may even be a low estimate depending on whose research
you use) for a such a company? I can't come up with a good reason,
although I can come up with several bad excuses. Does Goldman think
YOKU should be worth SUBSTANTIALLY more (less) simply because its listed
in the U.S. instead of China? Huh???
Zhu & Zhu (2010, pdf)
estimate Risk Free (Rf) and Equity Risk Premia for Shanghai, Shenzen,
and Hong Kong stocks at 11.23%, 14.04%, and 8.19%, respectively. Using
my beta and these rates (along with Zhu & Zhu's "adjusted ERP's"
which are significantly lower), we get the following values for YOKU's
Equity Cost of Capital using CAPM:
Because
the adjustments account for factors I don't think are relevant here
(e.g. limited investment opportunities faced by mainland Chinese
investors, etc), I prefer to use the unadjusted numbers. So doing
indicates Goldman's Equity Cost of Capital is anywhere from ~22%-42% too
low, which has a profound effect on the ultimate company valuation. To
give you an idea of just HOW profound the effect is without replicating
Goldman's numbers, if I change the Cost of Equity in my model from 15%
to 12%, the valuation changes from $21.37 to $38.31, an 80% increase!
If Goldman had used an appropriate Cost of Equity in their valuation,
their target price would be MUCH lower than $55, likely around the
stock's current price of ~$30. After reading a fair amount of research,
I just can't imagine any way GS can defend using 6.5% as the Cost of
Equity, at minimum, that number is ~150bps too low.
I'm not a huge
fan of using Beta to estimate the Equity Cost of Capital (there are
other, slightly less "scientific" ways of so doing), but what I think
this exercise shows is that Goldman's analysts have SEVERELY
underestimated YOKU's risk as measured by its Cost of Capital, which
results in a SEVERELY inflated valuation.
For what its worth, I'm
also not sure where Goldman got 1.2 for Beta, as my calculations using
the NASDAQ for the Rp show it to be 1.31 (~8.5% higher) , even if we
assume they only included prices up until the day before the report was
dated. I suppose they could have used a different index's returns for
Rp, or used weekly instead of daily returns, or something. Why Goldman
used 6.5% for the Cost of Debt is also a total mystery to me, though, as
the footnotes to the firm's SEC filings (pg F-24 in the 5/2011 prospectus) say the interest rate is 12%. These are (slightly) less of an issue than the grossly understated ERP though, so I'll let it slide for now.
If
I'm feeling adventurous, I might just contact Goldman and ask them to
explain (defend?) their numbers. I'm not going to lie, absent some
fantastically surprising explanation, it looks like Goldman just did a
goal-seek on the Equity Cost of Capital to get to the price they thought
the stock is "worth," which is a MAJOR screw-up the likes of which a
summer intern wouldn't even make!
Its one thing to claim revenue will
grow or margins will expand faster than other, less biased market
participants think to get your target price a few dollars higher. Its
another thing altogether to use such indefensible and wrong numbers to
jack up your valuation like this. I think no matter how you look
at this, it looks bad for Goldman.
At a minimum, Goldman has
produced a report with an inappropriate Cost of Capital
and didn't update it for country (etc) risk, causing The Market to
assign (insofar as GS reports are relied upon) a SEVERELY inflated value
to the stock. At worst, we're looking at a gross, intentional, and
possibly systematic abuse of valuation techniques in a report for a
stock Goldman underwrote, a situation which if true, I'm sure the SEC
and oodles of securities lawyers would LOVE to hear about...
- advertisements -
Further at the root of this is the typically American inclination to avoid direct experience, to believe in the interventionist hero (a broker), to not evaluate things for themselves because our technological revolution has allowed weak-minded, politically driven people to be involved with things like finance for which they have no real mathematical or ethical aptitude. So we have jerks like GS still shilling to their jerk clients. Why do we care about policing them? Why put more time into it? It reminds me of that dude who got attacked with a pole in the Vancouver riots: sure, the guy took a swing. He takes the pole out of his hand (!) but chases him into the crowd (?) where the crowd gangs up on him (!) and kicks his head in (!?!). The victim was defending a. storefront. Why was he defending a. storefront. Which is insured. Is he just a closet racist with his big chance to be a "civillian sheepdog"? A tea-partier with a pillowcase of guns waiting to come charging out of his Costco shed with those flimsy aluminum doors and a home-made Ghille uniform with branches sticking out of the helmet he got on Amazon behind his house to kill 6 people, only to get stormed by dozens more?
It's almost as fruitless as policing the likes of GS, and/or their investors.
GS' so-called "service" isn't worth the scrap price of the building in which they reside.
So the current mess involves a history of Goldman Sachs persuading bond investors to buy Greek bonds on near identical terms to German bonds; as though the Greek ones were also very low risk.
And they appear to be pulling the same stunt again; only somehow it got spotted.
At the root of this, is a belief in the work of quants as being scientific with absolutely no need to sanity check it against human nature, commonsense, wisdom or integrity. At the core of that belief is either immense naivity or a deeply cynical version of the greater fool theory. To repeat the mistake suggests it is the latter.
Why would we use the U.S. ERP (6.5%, which may even be a low estimate depending on whose research you use) for a such a company?
Why would the 5-Yr TBill for a structurally defaulted, insolvent country attract a single buyer with a 1.45% yield? Stuff's gotta come from the same corrupt wonk factory...
sadly everyone lies to congress...
"sadly everyone lies to congress" is an anagram for;
sadly congress lies to everyone
Sadly, everyone in congress lies.
I'm sure the SEC and oodles of securities lawyers would LOVE to hear about...
And that of course, is where you would be wrong.
Sadly, you are probably quite right.
Goldman's had better watch out. They'll end up with a $40,000 fine from the SEC.
LOL
Why such a big fine - does the SEC need to make its quota?