Vaguely condescending blog posts by founders of financial firms about why top-tier financial reporters are underestimating their returns is a leading indicator that actual returns will be even worse than the reporting implies.
"When the facts are against you, argue the law; when the law is against you argue the facts. When both are against you, call the other lawyer names."
For financial services the equivalent would be either talking your portfolio or your returns. In this case, the mark-to-market portfolio looks bad and the returns look bad, so they've resorted to arguing with newspapers.
Not necessarily. VCs' lawyers are big big sticklers for preserving their Reg D exemption from SEC registration, which is contingent upon no general solicitation.
If AH wants to be able to sell any LP interests in the next 12 months they are going to be very, very careful not to state any performance figures, particularly the trumpet-able kind.
(Exception: there do appear to be firms who are cavalier about this kind of thing, mainly seemingly new or nontraditional firms. But if you have DLA or Gunderson or Proskauer or whoever it is these days advising you, they are not going to be cool with you possibly blowing your Reg D in order to have a twitter feud with a reporter.)
no you won't, the VC doesn't report to a random reporter and numbers change all the time (WhatsApp last public valuation was of 1.5 billion (Sequoia held them at this position) and it sold for 19 billion)
That doesn't even make any sense. The average actual life of a VC fund is, according to numbers I got from NVCA a couple years back, over 16 years (despite being originally targeted at 8-10 years). If you wait until a fund is "closed," you are going to be waiting a long, long time.
The industry already has standard measures for talking about performance of a fund from different perspectives. I fear that HN is reinventing several wheels here.
For example, TVPI, Total Value to Paid-In, includes "marks" (which the VCs strongly influence, but ultimately have to be signed off on by auditors). But every savvy industry person would also look at DPI, Distributed to Paid-In, which includes only actual cash (and marketable securities) distributed to investors, and as such is much less susceptible to gaming.
It would be the scalar fallacy to believe that you can reliably compare any two funds' mid-life performance with a single metric. A high DPI fund is much more certainly a performer, but might have less residual value in the portfolio; likewise, a very high TVPI fund (but with low DPI) might have a ton of realizable value or it might just have unrealistic marks. (Finally, neither of those measures accounts for time value of money.)
So I agree with the sentiment that marks are not useful however you can do an analysis of the "non-realized gains" and the age of each fund then compare that to benchmarks of other funds of the same year.
That's basically a backend way of figuring out how much of their returns are pie in the sky. One other thing to keep in mind to give a16z the benefit of the doubt is their funds are MASSIVE. Putting that amount of capital to work is hard- and relatively harder than Bessemer and the other funds they are compared to.
That said, the % of Fund III in particular that is non realized is fairly scary. I'd take a wager that it does not return capital. Note also A16z pretty quickly raised a new fund before a lot of these returns for 3 become clearer. That's just guessing on my part but I'd def bet III is not gonna be good. That said I still love a16z and the partners and their approach...I actually hope I'm dead wrong.
> you can do an analysis of the "non-realized gains" and the age of each fund then compare that to benchmarks of other funds of the same year.
Just so I'm clear, the idea here is to look at all funds that started in the same year, average the non-realized gains, then compare that to the non-realized gains of the fund you're evaluating? How do you know whether the difference is how they mark vs. actual alpha?
Look at the funds a specific "vintage" year, then look at:
1) Realized Returns/ Total Size of Fund
2) compare above to prior vintage years of same partnership/ similar funds
3) Look at unrealized returns as a percentage for this vintage across funds and for same partnership for prior funds for same years out.
The bulk (like 80%) of unrealized gains can disappear in a quarter if you have a large market correction.
Some of the variance is going to just be market conditions but in general the weaker those numbers that's a pretty bad sign.
From the original WSJ article: "Since its founding through last year, Andreessen Horowitz had returned a total of $1.2 billion in cash to investors, net of fees. Sequoia returned more money on WhatsApp alone."
Unlike the blog states, it seems that WSJ is, in fact, using "actual, realized returns" as evidence.
VC as a fund class is literally about finding outliers. How much of that is luck is arguable* but it doesn't make any sense to ignore outliers when comparing returns.
* How did Sequoia end up as the only VC that made money from Whatsapp? It certainly wasn't an accident.
Great link thanks. I had no idea they were that effusive about venture money. Really hammers home the idea that if you have a killer tool, money will find you.
Jan and Brian were uninterested in VC or business relationships to the extent that they were impossible to contact. There was no address on their website. There were no contact details. They ignored inbound emails. There was no signage on the building. They ignored press emails. But Sequoia knew they were in Mountain View and literally had partners walking the streets doing a physical search for them in order to initiate that contact. That's a pretty amazing effort.
Additionally, Sequoia have an internal tool called Earlybird they wrote themselves that monitored mobile app stores, which is how they noticed that WhatsApp was doing so well despite the fact that it had not taken off in the USA specifically. They relied on hard data collected systematically, rather than waiting for hot Bay Area startups that the VC's friends were using to turn up on their door.
And finally Sequoia had a great name that Koum associated with success.
There's a lot of great info in this video. I love WhatsApp partly because it violates so much received valley wisdom and yet has been so successful.
I read the title and I thought "mark" as in "the target of a scam". I was super excited to read a post about how VCs scam founders but get away with it because they're, well, VCs, and on a VC blog, no less! Imagine my disappointment on clicking the link...
One thing that the article doesn't really elaborate on is why different valuation strategies exist (and are allowed to exist) within GAAP.
Accounting should be done in a way that maximizes the usefulness of financial reporting for strategic and management decisions. In some cases, firms are required by regulators to adopt specific practices, but there is a fair bit of freedom given to the CFO.
VC firms must make wise financial decisions and satisfy their LPs with some degree of transparency. The accounting strategy chosen must accomplish both goals.
In many cases, it makes sense to be very pessimistic about valuations, and doing so often reduces tax liabilities.
On a side note, the whole "mark to market" scandal from the 2008 financial crisis was a case where firms typically marked assets in a way that matched their management goals, but at times failed to reflect short-term price fluctuations.
Regulators thought that forcing firms to mark assets to a known market price would result in better financial reporting. The problem was that the balance sheets containing those assets were also used as underwriting capital. So a market price increase (or bubble) in the assets was suddenly leveraged into a lot more risk capital by the firm (whereas before the rule, the CFO would not likely have wanted to mark the assets that high).
This resulted in industry-wide increases in risk capital for "free" because of asset price spikes, and following that it led to increased investment. The problem was, when the price fell back down, the firms were over-leveraged. It's generally a bad idea to use highly volatile assets as underwriting capital.
So while "mark to market" sounds good, it can enhance natural fluctuations (minor boom/bust cycles) in a destabilizing way.
The art of being the CFO of a VC firm is likely a very interesting thing, and it would be fascinating to learn more about how this happens across the industry.
OPM is terrible. Even the academics realize it (I think). This paper evaluates the analysis. There's a caveat on pg. 24. GIGO. In my opinion, it's not just a caveat. It's a killer. I'm not sure what the overarching societal value to applying OPM to private, illiquid company valuations could be.
I don't buy it. They assign value when then invest (which of course is a very rough estimate). What's so hard about assigning a value a few months or years down they road when they presumably have a lot more info about the biz than they did on day 0 (the initial investment/mark).
Rare for a Finance Analyst at the WSJ to make such a simple mistake, but is it so surprising that a media outlet would overlook accounting details in the chase of a headline?
