This is the 'bubble' that people in the tech industry should really be afraid of, says the leader of Silicon Valley's top startup school

Sam Altman, the leader of startup school and venture capital fund Y-Combinator, doesn't think we are in a tech bubble.

He believes we are in a "tech bust," where technology companies are undervalued. He thinks private technology companies as a general group are destined to rise, and that people obsessing over individual valuations on individual companies are missing the big picture.

However, not everything is perfect in startup land, says Altman.

Altman was the guest on the Jay and Farhad Show, a weekly podcast I host with New York Times' columnist Farhad Manjoo.

On the show, Altman says, "The bubble I'm afraid of is the risk bubble. People are so sure that things are going to work out that you see burn rates that I have never seen before, and that I think is actually a real danger."

Right now, startups believe everything is going to work out so they are spending recklessly, burning millions of dollars every month. This will not end well for many of them.

You can listen to our full audio here below. We've also transcribed, and edited the conversation.

You can subscribe to the podcast in iTunes here. Or just look for it in your favorite podcast app under "Jay and Farhad". Here's an RSS link to the show. We use SoundCloud as a host, so you can listen to the show over there, too.

J &F: I've looked at your blog archive, and you've written about the economics of startups quite a bit. Do you think startups are not charging enough for their products? Are they spending too much money?

SA: Well, actually I think they are two different but sort of interrelated problems, so let me take a step at articulating both of them.

One is what I call the unit economics problem. This is a sort of newish thing where they lose money on every transaction and they have plans in the future for how they are going to make money later. I wasn't around for the 2000 bubble, but I've heard this was super common then. So you have a lot of either enterprise SaaS or on demand companies or whatever you want, where they are selling a product for less than it costs them to deliver it, and they have a plan - or maybe they don't have a plan - to make it better later. And look, it's true, if you sell a dollar bill for 90 cents you can grow really, really quickly until you run out of cash, and some of these companies do grow incredibly quickly. But I think it's dangerous to grow, and bad advice that these company sometimes get from their investors, which is just like, 'it doesn't matter how much money we lose per customer just keep growing and then we'll be able to raise more later,' and sometimes there is just no plan for how that ever gets resolved.

So I think that is one problem, where you have a company that has very bad unit economics, they lose money on each transaction, and they lose ever-growing amounts of money every month. And when you ask the founders how they're going to fix this, sometimes you get a little bit of a glassy stare.

A second issue is companies raising at very high valuations, because - or at least as I believe - that if you raise a lot of money in preferred stock with certain terms then valuation doesn't matter.

However, where I think people kind of get screwed with that sometimes, is no one ever wants to raise a down amount, for good reason. And employees, and this is a bad thing, I think employees see someone like Fidelity or whatever come in and price a company and say, "oh Fidelity's an expert, this company must really be worth $5 billion, and thus my options are worth x." And then they find out that that was a fiction, but the company doesn't wanna raise a down round and deal with all of that.

So you do sometimes get trapped by these high imaginary valuations. They are related because you need more money to lose more money, but they are two independent problems I think.

J &F: Recently, Bill Gurley said something to the effect that that startups have to play the game that is on the field, not the game as they want it to be. How do you reconcile a conservative approach to business with competing with rivals that might be willing to do crazy stuff?

SA: That's a great question, and it's why running a business is hard. There is no-one-size-fits all answer for that. It depends on how confident you are about your customer lifetime value, how confident you are you can raise prices, how cheap the capital is available to you, whether you believe the competitor is going to die or not. These are the conversations that take two hours in a board meeting, and you know if you have five people on a board, you have three or two different opinions that are very reasonable, and these are the hard judgment calls to make. It's true that sometimes a reckless competitor will beat you in the long run, but I don't think a general purpose good strategy is, "well my competitor is doing this crazy thing and lighting dollar bills on fire, so I better light more dollar bills on fire." That does work occasionally, but I don't think it's a general strategy that you can recommend without knowing the situation.

I do appreciate the point that Bill is making, that you have to play the game that's on the field. But I think he's also made the point that you shouldn't be totally reckless in business. You should not have crazy burn rates, so I'm more inclined to agree with him on that advice and maybe sometimes you shouldn't actually play the game. Maybe sometimes the game is just crazy.

I appreciate the comments Bill makes about not having crazy burn rates, but one company he's on the board of, Uber, has burnt more money than any private company ever, I think. So, obviously, as Bill is proof of, there are exceptions and sometimes you really do want to spend a lot of money.

J & F: In your latest post you say the public market valuations of tech companies are far from bubbly. Why do you think that is? Why are public companies given conservative valuations?

SA: You know I am so far from a public market expert that I would only be guessing to say why they decide value a company the way they do. I think the efficient market hypothesis is clearly wrong plenty of the time. Like, Amazon has doubled in value in the last 10 months, so clearly the efficient market hypothesis got something wrong there. I don't think they've transformed their business such that you can blame it on that. I think that I don't understand the way the public markets think about the going forward value of Apple in particular, for example. And so that's ok with me. I am still willing to say, "Well, I don't understand why it's mispriced, but I do believe it is mispriced, and that's why I'm going to keep buying shares."

J & F: Earlier you said you think that some startups are overly optimistic and taking on too much risk. You also said that you think some high valuations are irrelevant. How does that not mean things are in a tech bubble?

SA: It comes out to the sort of portfolio construction thing. I think if you buy a basket of public and private tech stocks, even with this crazy debt disconnect, you'll still make money. But on an individual company basis, I think there are particular companies that are going to kill themselves by spending money.

J & F: Do you wanna name some of them right now?

SA: (laughs) Not really.

J & F: Do you think there are gonna be a couple or a whole bunch that are going to be doomed?

SA: More than a couple. I don't know how to put a percentage on it.