Hacker News Reader

Comment by twblalock | Hacker News Reader

Every year since the current boom started, people have predicted that we are in a bubble that is about to burst. Every year since then, that prediction has been wrong.

It's pretty clear that most of these predictions have been motivated by sentiment or mood affiliation (e.g. "these new startups don't really make anything, techies are ruining San Francisco, so they must therefore be bound to fail") rather than economic facts.

So, I just don't listen to those predictions anymore. If a slowdown, or a slump, or a crash happens, I'm sure I'll notice it without the naysayers' help. I'll start to worry when it becomes significantly easier to hire engineers.

View on HN

I like all the constant bubble talk. As I see it, the more wary people are, the less likely we are to collectively make dumb decisions that lead to an actual bubble. It's not like any of this talk goes around with total certainty.

Making WebVan style bad investments because you can resell it is only part of the problem. The other part of the problem is systemic: it only takes a few dumb actors to make everyone act poorly. When AIG underwrote credit default insurance priced premiums too cheaply for the eventual outcome, a lot of people had no idea how widely distributed their mistake was -- because contracts weren't tradable, and a number of firms were engaged in contract type arbitrage, nobody had any clue what percentage of their liability traced back to AIG. But it wasn't zero just because they never bought from them -- you needed to know your transitive exposure to counterparties to properly offset risk.

Obviously AIG got to where it did because of a flawed system that allowed AAA rated companies to issue insurance with no capital, but the analogy of a system of largely rational actors making collective mistakes still holds. One collective bubble argument in tech is ad driven. Every ad driven app & game I've tried has been largely advertised other ad driven games. This represents and form of investment -- spend ad money now and get back more ad money over time. If a few actors are overly optimistic, they'll be outbidding other advertisers and raising the return on ads.

Which can produce a couple of adverse feedback loops. Firstly, other companies look at the going rate, and have no way of predicting a crash in ad rates. But if the rate is high now, they will probably choose to produce more ad driven apps. Second incumbent advertising networks (Google, Facebook) will derive immediate benefit, and have extra cash to invest to support more opportunities further out on the risk-reward spectrum. Thirdly, the optimistic actors themselves have a sort of feedback loop -- if rates rise based on their entrance, then their models may well predict rates continuing to rise. Emboldening to buy more, and higher.

The trouble is, we don't know if this is the case or not. It seems possible, but there's also a number of micropayment based games that might explain an increasing ad rate. And we can tell ourselves that ad blockers are way less common and less effective on mobile, so maybe we shouldn't expect desktop and mobile markets to converge to the same place.

VCs would knowingly fund a bubble if they thought they would still make money after it burst.

Similarly, before 2008, mortgage lenders knowingly lent to borrowers who were very likely to default, because the lenders knew they would be able to bundle and sell off the mortgages before they went bad.

Investors who know there is a bubble will not always sell right away -- they will wait until their investments grow even more, and then sell before the crash. Some will pull this off, and some won't.

So, knowledge that we are in a bubble won't always remove the incentives that created the bubble. That's why bubbles keep happening.

If the anything they aren't making is money then it is a bubble and it will eventually burst. When growth in users is seen as orthogonal to growth in income then the whole thing will come crashing down. If growth in users can be consistently converted to growth in income then it's not a bubble. Many people are skeptical of whether user growth can be consistently converted. Very true observation on the effects of a burst bubble -- if it happens everyone will notice. No one is good at predicting the end of a bubble so it is generally better not to pay attention to the predictions until you start observing effects directly.

I don't believe anyone who won't buy a bunch of options short the companies they describe as going to collapse. I mean, if you _know_ something is going to collapse, you've got a money-making opportunity. And if it's the total collapse you're predicting, you can make out like a bandit. If you don't have the money to do this, I don't really trust your ability to make money anyway.

That doesn't mean I believe anyone who backs their prediction with money. Necessary but not sufficient condition.

There are thousands saying it's going to the moon tomorrow, and thousands saying it's going to crater tomorrow. Someone's going to be right no matter what. When the cost of failure is zero, everyone's going to have an opinion and there's no way to know if anyone's opinion mirrors their true preference. Their true preference may yet be wrong, but at least it's their true preference, so it may be worthwhile to spend time finding out why and see if it truly applies.

The market can stay irrational longer than I can stay solvent. :)

All the more reason to not make predictions one can't back, then, isn't it? :P

Really the only retort to "put your money where your mouth is" talk.

I'd be interested to hear your take on what these economic facts are. Just curious.

Presumably this is highly weighted toward economic facts: https://fred.stlouisfed.org/graph/?g=6w6u

Reply on news.ycombinator.com