Hacker News Reader

Comment by cs702 | Hacker News Reader

Meh. The OP looks only at historical data only over the past 10 years -- all of which occurred during the longest-ever rising (bull) stock market in US history (excluding the temporary shock and recovery at the beginning of the pandemic).

A more useful analysis would look at data during the 2000 to 2002 period, during which the market caps of many prominent, high-quality tech stocks declined by 80% or more. What happened to startup fundraising during that period?

Or it would look at data from the 1973-1982 period, during which many so-called "nifty fifty" tech stocks (think IBM and Xerox in their heyday) declined by large double-digit percentages and did not recover for nine years. What happened to startup fundraising during that period?

View on HN

It's a fair point. I didn't go back that far in the data set. I think the one major difference between the last 10 years and the.com era was the lack of viable business models in that time. I think the crypto ecosystem looks more like 2000 than the current software market because most of the software companies today worth billions generate hundreds of millions of revenue. Also, the interest rates within the last 10 years are going to be most similar to the ones that the FED is contemplating implementing over the next couple of years. The rates in the early 2000s were significantly higher in the 3-6% range. https://www.macrotrends.net/2015/fed-funds-rate-historical-c.... Last, the third trend that's important is just the volume of dollars going into venture Capital which is 20 times what it was during that period of time

> lack of viable business models

To be fair this still remains to be seen for most? public tech companies. It’s easy to be unprofitable and just say you’re following the Amazon model. But can Uber or AirBnb actually run a business that increases their net profit YoY sustainably? I think that remains to be seen. The super easy money in both private and public markets for over a decade has kicked up this fog so that it’s hard to make out what is going on. It could be that many are just as sustainable in a tight market as they would’ve been in 1999

Of course no company can increase their net profit YoY sustainably, unless you mean increasing it at the rate of overall economic growth, because otherwise after a few decades the economy consists of only that company and its growth must then fall to the rate of overall economic growth.

So let's consider the more reasonable interpretation of "viable business models" as whether Uber or AirBnB can be sustainably profitable at a level that justifies their current market cap (and thus stock price.) Uber Cab is US$73B, AirBnB is US$100B, and justifying those at a P/E of 20 would require respectively US$3.5B and US$5.0B per year of profits, which are US$0.50 and US$0.75 per person per year, at the current population.

They're both effectively marketplaces making a living by skimming a commission off the earnings of actual service providers, who are providing respectively transportation and lodging. Such commissions are usually in the 0.5%-50% range depending on the market power of the marketplace and the kinds of risks involved. (Consulting agencies commonly suck off 50%, for example, in part because they don't get paid at all if the client is sufficiently unhappy.) Suppose they're 0.5%. Then to justify Uber's market cap, the annual transportation budget per person needs to be at least US$100, and the annual lodging budget per person needs to be at least US$150.

These seem like very plausible numbers to me, even a bit low. For example, right now Uber's commission is 20%, 40 times higher than the 0.5% I'm using above, and pre-covid the US hotel/motel industry was about US$220 billion for a country of only 330 million people, which is US$666 per person, a beastly number that is four times higher than the US$150 I computed above. So I think the market is pricing in a fairly large risk that Uber or AirBnB will just blow up and be worth $0 in a few years.

But I guess you see it very differently. What's your analysis? What am I missing?

Commissions are revenue, but it's earnings (profit) that you should really care about. The mentality of "Revenue Growth At Any Cost" (spend $1.2 in CaC to buy $1 of revenue) is coming into an end, and my view is that only profitable businesses will survive.

Airbnb only reported positive earnings in the last quarter (which is great, hope they continue the trend), but I reckon a lot of "never made money" tech companies are going to continue to see their valuations sinking.

Absolutely true, and I should not have conflated revenue with profits in my comment. Plausibly AirBnB and Uber won't have a significant cost of sales in the long run, but plausibly they will. I doubt they'll have retail-like 5% profit margins though, so the difference might be a factor of 2 or 3.

Isn’t this line of reasoning assuming Uber / AirBnB capture 100% of the market? I don’t think that’s a realistic assumption.

Yes, more or less. AirBnB does have pretty close to 100% of some overnight lodging markets and are likely to take more, but they're a long way from having 100% of the market now; their revenues are only about US$5 billion a year.

But I think the major threat to their business is not competition from the old guard with their trillion-dollar yearly revenues; they'll eat those guys' lunch, no problem. The real risk is that they'll be banned, like they are in New York, or soft-banned to favor local startups, like they are in China.

> Then to justify Uber's market cap, the annual transportation budget per person needs to be at least US$100, and the annual lodging budget per person needs to be at least US$150.

This is also ignoring potential for expansion into other markets. For example, Uber's model might be useful for trucking services. You bought some used furniture, you want a truck to come and bring it to your house. You're a plumber, you need a water heater transported from the warehouse right away for an emergency job. Open the app.

Uber has a large number of drivers connected to them, why not go into competition with UPS and FedEx?

Maybe things like that succeed, maybe they don't. But if they do they make a lot more money, and the possibility that they do makes the company more valuable.

They've already launched Uber Eats here, going into competition with PedidosYa, Glovo, and pizza delivery boys. And they have a package courier service which was actually their only service for a while at the beginning of the 9-month lockdown almost two years ago. So, sure, they could expand into other markets. But I don't think that's what investors are betting on.

This market repricing can only be temporary (at least in nominal terms), because where else would money go? For decades interest rates have been shrinking, making fixed income a safe investment. That’s not the case anymore, interest rates are on the up, where should money flowing out of the stock market go? You are going to see more sector rotation within the stock market, rather than money flowing out of it.

Today’s world depends on tech to an incredibly high (and increasing) extent. Eventually that will be reflected in market valuations.

The venture capital industry today has a much more mature and robust structure than even 5 years ago: the structure of the funds, the fund commitments, the size of the funds and the management companies. They are much more resilient to market moves than in 2000.

Thanks!

> I think the one major difference between the last 10 years and the.com era was the lack of viable business models in that time

In quite a few cases, yes -- I agree. In many other cases, though, I think it was the sudden unavailability of capital that did in some companies that otherwise could have been viable. Given enough time and money, poster-child dot-com failures like Pets.com, WebVan, and Kozmo.com might have well become a Chewy, Instacart, and Doordash.

> Also, the interest rates within the last 10 years are going to be most similar to the ones that the FED is contemplating implementing over the next couple of years. The rates in the early 2000s were significantly higher in the 3-6% range.

Maybe. I for one wouldn't dare make any predictions as to what will happen to rates. I mean, in 1972, when the 10-year treasury was ~5%, no one -- no one -- even imagined that it would increase persistently for a decade until it ~16% in 1982. In 2000, when the 10-year rate was ~6%, no one -- no one -- even imagined it would decline persistently over the next two decades, hitting ~0% in 2020. Today, no one -- including me -- can imagine how or why rates would go all the way up to 6%, let alone 16%... but I'm sure the future will find a way to surprise us.

> Last, the third trend that's important is just the volume of dollars going into venture Capital which is 20 times what it was during that period of time

It depends on whether that large flow would be impacted in the face a persistent bear market -- e.g., could a large proportion of LPs default on capital calls due to losses in other markets? Keep in mind, VC funding declined by a factor of ~10x in 2000-2002. Could it decline by a factor of 10x this time around? I think that's unlikely, but we can't rule it out simply because the amounts are larger.

You’ve got to include market corrections or some model of “down” as well as “up” in your data or your model will only be optimistic. Your model can’t see past what data it has.

> "the crypto ecosystem looks more like 2000 than the current software market because most of the software companies today worth billions generate hundreds of millions of revenue"

The big difference between the dot-com bubble and the current cryptocurrency bubble is that there were lots of useful dot-coms which didn't make a profit while there are lots of profitable cryptocurrencies which aren't useful. That's because dot-coms put the use case first while cryptocurrency puts the monetisation first, leading to the monetisation of nothing and removing the incentive to actually build anything (whether useful or not). It remains to be seen whether any useful legitimate practical uses will come out of the cryptocurrency space, but it isn't looking promising after 13 years, while after that length of time the dot-com bubble had been and gone and recovered and led to mass adoption of lots of useful things we all use on a daily basis.

On the point of "useful legitimate practical use", I can send or receive money cheaply and easily anywhere in the world without any type of account or identity verification.

That's pretty cool.

Agree on the rest. It's overinvested.

In all seriousness: I looked into moving money from Canada to the US. It seemed like between gas fees and exchange buy/sell spreads that I was looking at an effective cost of 5-10% at a minimum, higher if I wanted to avoid KYC rules by transacting off the exchanges. And that is ignoring any volatility. That’s about double or triple what I would pay using a traditional bank.

Can anyone point to a practical way of moving money?

PayPal? It’s free and instant and has been for decades.

Haha.

I meant in the context of international transfers between countries with different currencies which is one of the cited "success stories" for crypto. My research is that it is actually a pretty bad way for me to spend my money across boarders. I haven't seen a practical, affordable, way to transfer money via crypto that is better in any way than retail banking. I suspect that outside of Iran and Venezuela, crypto is pretty useless as an international commerce device.

Have you heard of Transferwise? Would solve your issue.

Separately, if I send you some Bitcoin right now, the fees will be tiny, not sure how you're getting to 5-10%. Volatility is a real concern and you can easily move 5-10% because of that.

ETH is a different story.

Launched my career late 1999. I was working a lot of short gigs to build experience.

Multiple calls from recruiters daily. Then summer of 2020 the calls stopped. Six months later I gave up started at RadioShack. Where I met engineers with10-20 years of experience that couldn’t find work.

I take it you mean summer of 2000?

Yes

Something that doesn't really work or make money today, like all the crypto thing, might probably follow the same path. But tech is now, unlike in 2000 or 1973, is holding society by the throat, it is hated as much as oil companies were hated in 1973. So i don't think they are going to collapse nearly as badly.

Agree with you, but a semantic point of correction is that, while the nifty fifty did include many emerging tech stocks AND were mostly focused on growth industries, the index included lots of non-tech, like JCPenny, McDonalds, Sears, Walmart, Gillette, and many others.

It would also be a lot more compelling to show number of deals. I get the impression it's deal quantity that slows. Size is held as more of a standard.

Or the window around the 1987 Black Monday/Tuesday event that turned the semiconductor industry into liquid turd and forced the creation of new trading "circuit breakers" to prevent it happening so glibly again.

Reply on news.ycombinator.com