Hacker News Reader

Comment by katmannthree | Hacker News Reader

From the news articles I've seen I get the impression that aside from an early investment in Alibaba, SoftBank has a very poor track record. Is that actually true or typical for VCs, or is it just a bad impression from only knowing about them from headlines?

View on HN

Sometimes VCs are a bit like hedge fund traders. They have a lot of terrible investments but one great one that made them some money and made more people want to invest. Often though there is little indication the initial one success was much more than just luck and they then proceed to lose a bunch of money for the investors that came in based on their “great track record.”

There are of course those that truly do have a great record over a long period of time but they are much less common. We’re coming off a period where VCs and others could probably have thrown darts at the wall and made a half decent return. Now that’s all changed so we’ll see who gets washed out and who can stand on their results long term.

As I think it was Warren Buffet once said, when the tide goes out is when you get to see who wasn’t wearing any pants.

As a bank you could probably hire 30 analysts who make random predictions over a few years. One of them would probably be right and you have your superstar who can attract clients.

Yeah, there's actually an old sports betting scam that works similarly. You send out letters to tons of potential gamblers saying something to the effect of 'The <team>'s games are fixed and I've got the inside info. To prove it, I'll predict the winner of the first n games' Then each week, you send half the group one prediction and the other half the other. Whichever half was wrong, you ignore, and the the 'correct half' you repeat with the next week. After n weeks, you have 1/(2^n) of your original group who've seen you be right each week, and at your desired n, you ask for money to keep going.

The modern version of that is writing a file for each possible outcome to a private Dropbox, and then delete all the wrong files. You can then make the folder public and tout how you clearly wrote it before the tournament began, because Dropbox wouldn’t lie just for you.

It's an interesting scam, but we all know how quickly n grows in this scenario. You'd need 66k potentials to have one correct guess, per team, per season in the NFL. And the "winner" of you lottery needs to have enough scratch to make your scam worthwhile, and be dumb enough to part with it for a peak of your post-season picks.

Why would you send more than 3 or 4 free "predictions"?

Yeah, not only you get a bigger pool of victims with a few predictions, and the idea is that you want there to be games left in the season, so they pay you money for the predictions of the rest of the season so they can gamble.

I think with some common sense and understanding of the sport you can probably narrow down things quite a bit.

Point spreads negate common sense and understanding. Remember touts sell you who to bet on, not who is going to win.

Turning to baseball there are teams like the '62 Mets that really help a fella out when "narrowing." The 2003 Tigers were not far ahead.

Yeah they made an episode of Alfred Hitchcock Hour based on that scam. So it's a golden oldie.

There is an old investment scam where you start with sending a few thousand letters with half saying gold is going up and the other half saying gold is downing down.

Initially, half the people got "good" investment advice. Send another letter to the people who got the "good" advice say the US dollar is going up or down spilt equally. Again, half the people got "good" advice, so send another letter to the people who got two lots of good advice ie a quarter of the original number. Continue to send letters to the "winners" with half saying buy a commodity and the other half to sell the same commodity.

If you started with 2,000 marks after four letters you have 250 people who would have made "a very large ROI" if you had followed our investment advice. Then you sell them some very bad shit or a subscription to your investment newsletter.

Morral: Past performance does not mean it will continue.

This can be done even easier these days. No need to target people. Just start a bunch of twitter accounts, then delete/hide the ones that are wrong. Promote the "correct" one.

I've seen magicians do similar things with prediction tricks.

If you are making lots of trades, you probably will need more than 30 analysts. Something close to infinity.

What hedge funds are you referring to? I ask because in my experience, most hedge funds take a much different approach than VCs. While VCs bet on low probability outcomes with potentially massive payouts, hedge funds usually bet on higher probability outcomes with low payouts. There are some exceptions, but in general hedge funds aim for a consistent return stream. As the saying goes, "bond-like volatility with equity-like returns."

That doesn't happen in hedge funds. Most large funds fire traders after the first risk limit is hit. When you are graded monthly, there is no room for error.

Most of the grousing about hedge funds is because they failed to keep with the market...guess who had the last laugh? I know one fund in particular that was huge, had big outflows because they apparently were out of touch...up 20% this year. Another one converted to a family office...they have been up 50%+ every year. There is also a big difference between the big macro, multi-strategy funds and the equity funds that are basically run with no regard for risk.

I'm not sure I get that impression from their list of investments -- check it out here [1].

The impression I get is that they are into extreme betting so for all their good investments (and they have many good picks), the gains get canceled by a few extreme bets on bad ones. I suppose that qualifies as "bad at investing" by some definitions though to me it feels more like "bad at hedging risk".

[1] https://en.wikipedia.org/wiki/SoftBank_Group

VC investment is characterized by frequent losses with the occasional outsized return. From what I've seen, they have a pretty good record with this frame in mind. How do they compare, peer-vs-peer? It's almost impossible to say since so few VC funds are public and because accounting presentation standards can vary wildly.

>pretty good record with this frame in mind

No, they don't. The average VC does not even return the S&P let alone make significate returns.

You always hear about funds making huge profits but that is for one particular fund for one particular VC. Even the best of them struggle to match 10-year index funds due to their cost structure 2+20%.

Large fund make heaps of money for the GP but not for the investors.

> From what I've seen, they have a pretty good record with this frame in mind.

I've always wondered if this is true over the industry. Obviously the top line VC firms perform quite well, but I'd imagine there is huge survivorship bias here (much like the rest of the actively managed investment industry).

VCs is the worst asset class - returns do not compensate for lack of liquidity.

As I mentioned earlier, purchasing ARM was definitely a good deal.

I think uber investment was also working out for them before the coronavirus hit (they paid $10B for 15% stake) . Apart from that I can think of Alibaba, ARM, Supercell, Flipkart that have been successful.

At least I remember them buying and quickly selling SuperCell off for a hefty profit. But that could be partly just luck. I feel they have had sound investment plans but they've also become victims of their own hype. Hype without reality is just delusion. And once you get engulfed with the feeling of self-satisfaction it is difficult to get rid of it.

They did great with Yahoo Japan. Also they pressured the government into guaranteed solar energy purchase rates and they built huge solar farms after the 3/11 Fukushima event.

How did their Sprint investment eventually turn out?

I read an article a few years ago where it was explained that investment is a winner-takes-all market: the top 1% VC firms get 99% of the best companies to invest in, while the rest is left with scraps (numbers as an example, not real). I am pretty sure SoftBank is not in the 1%, no matter how much money they have.

Wework could become huge as more people switch to remote work, I think a lot of people are missing the long-term bet here.

Yeah especially when a recession hits and people lose their jobs so they can finally start freelancing from an expensive desk and pay rent twice instead of working from home while they are struggling to pay rent... oh wait

a recession is not the issue imo. A recession is an issue for pretty much any business for a few exceptions.

The issue is that right now you have to stay at home. But in these rare confinement cases, very few businesses are valuable as well.

So I don't think that's a good marker. If at all, it's a good moment to buy wework.

They bought ARM. Quite a good investment I believe.

Reply on news.ycombinator.com