Hacker News Reader

Comment by HarryHirsch | Hacker News Reader

The extreme short-termism of capitalism that stifles innovation is missing from the analysis. Right now we are seeing an overinvestment in Sillicon Valley and an underinvestment in Biotech. The money is flowing to ad-selling services like Google and Facebook, and to intermediaries like Uber and friends, but biotech startups are funded below replacement rates. The development pipeline is drying up, it's not that the IP wasn't there, it's just not getting funded. Locust investors like Valeant don't help either. (Their business model is to buy up companies and then immediately divest of R&D.)

Any money in biotech goes to cancer instead of antibacterials, because the price points of cancer medicines are highest. The future was in the news last week: superbugs spread through colonoscopy procedures in LA. We'll see more of that in the coming years, but there is no alternative funding model within sight.

View on HN

Investment in biotech is low because the returns are low, not because of a high discount rate.

It's getting harder to discover new drugs and get them through the FDA gauntlet. Most of the low hanging fruit has been found, and additional fruit won't necessarily pass comparative effectiveness without huge (and expensive) trials. After it makes it to market you are subject to political risks - price controls, lawsuits over side effects too small to find in clinical trials, etc.

It's hardly clear that the valley is even a good short term investment. Most tech companies focus heavily on growth, not profits.

Why do you believe that biotech has higher long term returns than the valley?

Why do you believe that biotech has higher long term returns than the valley?

That's precisely the problem. Within my lifetime the current antibiotics will become useless, we will be where we were before 1937 (discovery of sulfanilamide), physicians will be as good as they always were at prognosis, but helpless in the face of bacterial infections. The apocalypse is coming, and capitalism is no tool against it.

The issue of fewer drug targets than in the past is not something that any economic system can solve. That's just chemistry.

As for the rest of it, it's true that capitalism can't fix the FDA or foreign price controls. Lets hope Charlie Stross is right and the capitalists can take over the government and fix this.

So, that's being too nice--recall that market forces are what have led to the overprescription of these medications and also the creation of the environments (hospitals) in which they flourish and mutate.

Yes, we would likely be better off with a Pigouvian tax on antibiotics - that's the economist solution. Unfortunately, neither the capitalistic system nor the democratic process has any real way to get us there.

Pigouvian tax on antibiotics - that's the economist solution

The Swedes and British manage extremely well with national guidelines that physicians actually adhere too. In some disciplines there is no alternatives to the technocrats.

Heh, the takeaway I've gotten from some of Stross's novels is that AI makes "capitalism" indistinguishable, at least from the point of human beings, from a centrally-planned managed economy. So there could be a future in which AI's become sufficiently advanced at serving human needs that we all live as komrades in a utopian communist society.

"Lets hope Charlie Stross is right and the capitalists can take over the government and fix this."

> ...but [capitalism does not require democracy] means that the interests of the public (labour) are ignored by states increasingly dominated by capital (because of [automation privileges capital over labour]) unless there's a threat of civil disorder. So states are tooling up for large-scale civil unrest.

Let us hope, indeed. Sounds like fun.

In the Chinese curse sense?

I don't understand how this is a critique of capitalism. Doesn't this argument of shortsightedness apply to all economic and political systems?

The specific defect here is that right now, today, a capitalist country is not, in your estimation, putting enough capital behind biotech to solve the problem of antibiotic-resistant bacterial infections. More expansively, you're saying that Capitalism is incapable of making this effort.

It sounds facile, but what's the alternative? Isn't this a defect of every economic and political system on the planet, not just capitalism? Cuban communism is not putting sufficient capital in play to solve this problem as well, you know. Nor is passive anarchy in eastern Africa putting enough capital towards solving the problem.

Even assuming that more capital would magically solve this is kind of a naive viewpoint; serendipity and advancement in parallel fields has played a much larger role in advancements in medical technology than directed research has [citation needed], and for all we know, advances in information technology could hold the answers to solving this problem in a much more effective way.

The notion that a directed, centrally planned effort can solve problems (rather than just implement those solutions) is one that is historically unsupportable in any context.

Within my lifetime the current antibiotics will become useless

Also, bananas will be extinct by 2008, and there won't be enough oil or food.

"additional fruit won't necessarily pass comparative effectiveness without huge (and expensive) trials"

Ah, that may explain the statistic I recently saw that Phase III trials are taking up 90% of the budget necessary to bring a new drug to market.

Think of a study that needs 3 lab test per day for each of the participants. These tests aren't cheap and a multi-year study with a couple of hundred participants with shipping costs (overnight) for the samples. It does ring up quite a bill.

Plus, oh, finding and keeping enough people in the study willing to get 1-3 blood draws a day???

Yow. I thought that by Phase III safety was pretty well established and you'd be looking for efficacy endpoints, surrogate or real. Emphasis on endpoint ... but I guess not.

But ... if vague memory serves, from the overview of some Phase III trials, maybe not the first one, I remember more like hundreds of participants. Maybe not quite a thousand, but a lot.

They don't have to be blood samples. I remember one which was blood, urine, and stool each day. Also, each sample might have multiple tests performed. Its a pretty interesting logistical problem.

> Investment in biotech is low because the returns are low

As I understand, investment in biotech is low because the time to realize returns is long, not because the returns are low (even after adjusting for time). The time it takes to bring a product from concept to market in biotech is virtually an eternity by Silicon Valley standards.

Also, the long term cash outlay for a drug that might have side effects discovered long after the patents run out. It is a very chancy business with a lot of failures per success. There are much better things to investment in.

The market's horizon depends on the world's population time preference and risk aversion as a whole, not on the whims of venture capitalist.

Why? Because arbitrage. If the world as a whole has a low rate of time preference and risk aversion, investments will flow towards high potential long term projects rather than short term ones.

But what if you're a greedy, myopic VC with high time preference who wants to make money now? That doesn't affect too much how you will value deals. What you care about is how much money you will make at the exit, which will depend on the equity valuation at that point. And that valuation, depends on the global investment's community risk aversion and risk preference.

Facebook trades at a P/E ratio of 71, meaning the equity is worth 71 years of current earnings. Netflix has a P/E of 116. Think that's shorts-sighted?

Ultimately, investments come from two sources, money saved and invested by individuals, and banks.

If it were only the first, you could say that the type of investments you see reflect the society you have. As a whole, society picks taxi cabs now rather than maybe some pharmaceutical product later. It's not capitalism, it's human nature.

Banks complicate the picture, since they turn their deposits into loans of various maturity. The rates on these loans are ultimately tied to central bank policies.

However, these policies generally attempt to keep rates low, so as to stimulate "productive" investments. In addition, various bail-out guarantees tend to artificially increase the amount of risk taking.

So there you have it. Capitalism by itself merely represents the time preference and risk preference of the population as a whole, and central banks tend to make it more risk taking and longer terms than investors would actually prefer.

> And that valuation, depends on the global investment's community risk aversion and risk preference. > > Facebook trades at a P/E ratio of 71, meaning the equity is worth 71 years of current earnings. Netflix has a P/E of 116. Think that's shorts-sighted?

Most humans aren't trading on earnings, but on market psychology, believing they can sell to a greater fool in the future. And most trades aren't even made by humans, but by HFT algorithms trying to win massively-leveraged short-term bets for a fraction of a percent. The Facebook and Netflix P/E ratios have very little to do with long-term earnings.

The only place you can criticize the (weak) efficient market hypothesis from is atop a pile of money. Where is it?

> It's not capitalism, it's human nature.

I disagree. Capitalism makes investments much more fungible than they would otherwise be, changing the incentive landscape. It can set up a prisoner's dilemma (i.e. a failure of global optimization due to induced local optimization) which preferentially allocates capital away from investments which are intrinsically difficult to value in a timely fashion or intrinsically difficult to capture value from (i.e. science).

Correct tulip speculation gets you $$$, making a correct incremental step towards curing cancer gets you nothing on the market. Nobody will be able to tell if your contribution was valuable for the next 30 years, and by then you won't have any mechanism by which you could retroactively hold back your contribution in order to obtain a fair market price for it. According to capitalism, this unfortunate (but completely incidental!) state of affairs is a good reason why investors should speculate in tulips rather than fund biotech. Yuck!

Can we do better? In the narrow sense of solving this particular problem, the answer is yes: a prestige-based economy (see: academia) is able to reward investments whose value can't be captured by a market economy, such as the "incremental progress towards cancer cure" example above. Of course, a prestige-based economy would fall flat on its face if one tried to generalize it to more mundane facets of everyday life, so it's not a satisfying solution for the general problem. But it IS a good way to get free-market drum-beaters to acknowledge their very constrictive assumptions about timely valuation and value capture :-)

That's a fair criticism.

You point out that capital will not flow towards investments who are difficult to value, or investments which are hard to extract value from.

I don't think the first point has merit. Any new venture is intrinsically hard to value, and even a global optimizer would be reluctant to invest in those. This is no different from risk aversion, which I've already cover. The second point on the other hand has merit. If it is difficult to extract value from an endeavor, you have an externality problem. However, this is a distinct concern than an alleged short-term bias in capitalism, an idea which I was attempting to refute.

The problem you describe exists just as much in the short run as in the long run.

As for what it is to be done, I think several things can help:

- Relax antritrust legislation: large companies who aren't threatened by competitors can capture a lot of externalities and tend to develop new science and technology. General Electric Research Laboratory, Bell Labs, Google X...

- Develop technology which lower the cost of coordination. Prisoners dilemma arise when there can be no coordination. In fact, if players can enter into arbitrary self-enforcing contracts, every game admits a strong Nash-Equilibrium, and the equilibrium is Pareto optimal! In practice, this means crowdfunding, and mechanisms like dominant insurance contracts[1]. Smart-contracts on blockchains are also interesting in that they are pretty close to the ideal "arbitrary self-enforcing contracts".

[1] mason.gmu.edu/~atabarro/PrivateProvision.pdf

> As for what it is to be done, I think several things can help:

> - Relax antitrust legislation

Why do you think it is better to fund labs through monopoly rent (where shareholders, administration, and inefficiency due to lack of competition all take a cut) than by simply increasing the budget funneled into the brutally competitive NSF grant process? If you assume that private labs are 10x as efficient as public labs, maybe, but I'm not convinced of 1x let alone 10x.

> Any new venture is intrinsically hard to value, and even a global optimizer would be reluctant to invest in those.

> ...

> mason.gmu.edu/~atabarro/PrivateProvision.pdf

So... crowdfund science by asking people to compute the monetary value of an "unknown unknown" (which is what science produces)? I don't think you would have trouble convincing people that the aggregate returns on science are pretty great, but I do think you would have a hard time getting the consumer excess of one of these contracts to exceed the friction induced from reasoning about unknown unknowns. A very hard time. Hell, I'd be willing to bet you that scientists figure out the secret to eternal life before bean-counters figure out how to crowdfund science at NSF scale with rational contracts.

Because even large trusts are still competing with each other. If the funding process within an organisation becomes too political, it will shrink. In general, a large corporate entity has good governance from the mere fact that it succeeded, it's good at allocating money to different projects and managing people. Governments are good at holding political power by playing through political alliance. Thus, I think they will tend to do a worse job in allocating grants. The process is likely to become more politicized than it would in a private organisation. There are of course exceptions, I am talking about general tendencies.

No matter who funds science, someone, even if it is the government, has to reason about the unknown unknowns when deciding what to fund and to which extent. The amount spent by the NSF is about $7B, which is a blip on the budget, and a blip in the amount contributed to private charities. The YMCA alone gets about $5B. The problem is that people just don't care that much about scientific research, and capitalism is good at producing what people want.

Or make people want what they produce...

> The extreme short-termism of capitalism that stifles innovation is missing from the analysis

What would that add? Civilisations don't have to innovate. In fact, the people on top may be perfectly happy with stagnation and the status quo.

Stifles innovation compared to what? I don't think there's any system that stimulates innovation more than capitalism.

Of course there isn't. Capitalism guarantees no such system is allowed to exist.

But capitalism is based on jam today and famine tomorrow.

This works just fine for people - as long as they're not thinking straight.

Intelligent resource allocation understands that famine ten years from now is a bad thing and should be avoided, even if it means lower short-term profits.

Any system that thinks ahead like that would be much more innovative than capitalism, because it would be able to make long term plans and invest in long term projects with pay-off times measured in decades or even centuries - something capitalism is manifestly unable to do.

It would also be better at avoiding the usual capitalist kooky idiocies like poisoned water tables, avoidable drought, accidental anti-terraforming, and the rest.

Public funding of science and universities stimulates more innovation. The stuff you see from companies took years of research in universities before they pushed it the last few yards.

"underinvestment in Biotech"

have you looked at biotech IPO's as of late. I would say the complete opposite.

It seems detrimental that large scale organizations are short sighted to 4 month (quarterly), annual, and 2-4 year (election) cycles.

Asphalt is cheap yet requires high maintenance and degrades quickly - rather than a larger initial investment in concrete which lasts 4 times as long.

Reply on news.ycombinator.com