What good is Wall Street? | Hacker News Reader

What good is Wall Street?

newyorker.com

90 pointsphilk61 comments

A well written, balanced article. As someone who has spent many years in the banking/investment industries on both the "buy" and "sell" sides I can say that there is no doubt in my mind that banks are an essential part of a healthy economy. I can also say without any doubt that banks have been allowed to go way too far and that it is all about the short-term rewards.

The real problem is not that firms took innovated and took risks, it's that no one seems to have been held responsible. All the "failed" banks, investment houses, and businesses should have been allowed to fail. They should have been liquidated to pay their creditors and their assets allowed to be bought out by firms that acted responsibly. Instead, the irresponsible were/are bailed out, and their managers are allowed to continue operating.

The lesson is: Take risks. You will pocket profits right away and bailouts will cover any downside.

exactly. The money quote: “In most industries, when people are paid too much their firms go bankrupt, and they are no longer paid too much,” he told me. “The exception is when people are paid too much and their firms don’t go broke. That is the finance industry.”

I'd like to make an amendment: Take risks. You will pocket profits right away and bailout will cover any downside as long as you can make enough people think you are too important to fail.

Another way to get bailed out is to make sure that folks who have the power to do a bail out think that keeping you alive will benefit them. You can do that with political support.

I was going to say something similar: Take huge frikkin risks that can sink the world if they fail. You'll make billions and bailouts will cover any downside.

I think a second big factor in this is that people tend to suffer from scope insensitivity, so it feels much nicer to win $5 ten times and then lose $100 once than it feels to lose $5 ten times and win $100 once.

I don't think that's true when you have incentives so that you make some proportion of the $5 each time you win, and don't have to pay any of the $100 when you lose.

I don't agree that the banks should have been allowed to fail in the heat of the crisis. That would have meant that our access to cash would have stopped.

I certainly get the point that, now that things are stable, the governement needs to haul all these failed bank leaders into a room and make them pay back the bonuses earned leading up to the crisis.

That's a nice, plausible position to take, and very popular too. Well done for saying what the people want to hear.

Now, taking a more realistic view, when Lehman collapsed, the stock market went into free-fall. The links between all those financial corporations are so thick that if one of them is allowed to fail, it threatens to bring the whole thing down.

If the governments of the world had not, with uncommon speed and purpose, stepped in and "bailed out the banks", money would, at this point, be more useful as fuel in a stove than as a bartering device.

The idea of letting failing companies fail is valid as long as they don't bring down the whole system. If you let companies grow to such a level of interconnectedness and size that they have the potential to destroy the whole world economy in a matter of months, and then you suddenly realise that they are that big and they are failing, the correct course of action is not to let them fail, but to sort out the short-term problem (i.e. bail them out) and then work on the long-term problem (cut down their size and complexity so they can once again fail without destroying everything).

Until a proper regulatory framework is in place to prevent financial companies from getting so intertwined and large that they can bring down the whole house of cards (for that's what every economy is, a house of cards based upon the common trust that it's all going to work out ok), it is in fact very reasonable to suggest that a good medium-term step is to get those companies to "stop innovating", given that their incentives are such that they will inevitably bring the whole thing down if allowed.

This is a more realistic view than the "let them fail" silliness.

You're ignoring the possibility of taking them over through bankruptcy and covering their potentially system-killing commitments directly. Bailing them out just puts more money in the hands of the people who created the problem in the first place.

+1.

Paul Krugman of the NY Times wrote extensively both during the crisis and since about the correct way to handle this situation:

1. Put the failing bank into "receivership". I.e., a "receiver" (temporary managerment) appointed by the government takes over the company and tries to salvage it.

2. Fire the original management team. I.e., the ones that got them into this mess with their excessive risktaking. Good. They should suffer the consequences of their bad management decisions. That's capitalism.

3. Take on whatever additional debt or equity financing is necessary to turn the company around - without paying any regard to the positions or desires of the current stockholders. The current stockholders' positions' will likely get diluted to nearly worthless. Good. They should suffer the consequences of their bad investing decisions. That's capitalism.

4. After the company is turned around, take it out from under government receivership, hire a new management, and (if needed/appropriate) launch it public again via an IPO.

There's a long, established history for handling failed banks this way (read up on the S&L Crisis), and it's the proper way to keep an institution alive and out of bankruptcy, without rewarding the management and stockholders who brought it to bankruptcy in the first place.

Krugman often pointed out, though, that the Obama administration was too scared to go this route - perhaps because they were worried about the absurd accusations of "socialism" that the Psychotic Right would undoubtably have started screaming about.

So instead, they chose to reward the management and shareholders that caused the problem, by using government money to help prop up the value in their worthless company. Moral hazard indeed! Privatize all the profits, but give all the losses to the public. And socialism is the evil here?!?!?!?

Your comment loses value through the unmitigated bias showing through. You had me until 'Psychotic Right' as though viewing the problems through a left/right filter is the explanation.

The fact is we'll never know what would have happened if large companies were allowed to fail, because they weren't. It might have been worse, it might have been better. For sure it would have been worse in the short term, but these issues should be measured over the medium to long term. One of the messages that has been sent out with the current handling is that there is a Fed put on any high risk adventure, and that's not good.

Personally I think that there is something to be said for the 'bandaid' approach. Not sticking something over the top to cover the problems, but a short sharp shock of pulling it off and let the problems be.

The problem with the bailouts is that they cover over the fact that some companies were allowed to fail. Lehman, for one. And the world didn't stop turning. Sure, this caused liquidity problems and risk premiums to escalate dramatically, to the detriment of employment and the real economy. But it's foolish to pretend that this isn't going to happen anyway. The problem with Krugman and others is that they believe a lack of demand is the problem. They view the problem through this prism, and as such they come up with solutions that match their world view.

Others (including me) say that perhaps the problem is excessive debt. And until that excessive debt is eliminated from the system, productive companies are unable to make use of the resources tied up in the unproductive ones currently taking shelter with public money. You say the institutions should be kept alive, I say that they should be destroyed and the buildings and staff and infrastructure released to new institutions who can find a way to be competitive.

I believe you shouldn't look at the problem in terms of a company or bank failing. That's just the symptom. Once the debt had built up excessively, it was only a matter of time and method for the companies to fail. The problem was watering down of regulation that had stood the test of time for 70 years, and regulators becoming toothless. So far, this hasn't been fixed.

This is a not a left/right blue/red debate. It is a debate as to the level that central governments should take in directing and manipulating markets. Neither 'side' is debating this. Nobody will try and defend central planning and a command economy, because it is comprehensively proven as a failed model, yet we seem to be reversing backwards into this situation on the say-so of a bunch of people who couldn't see the problems headed their way. They don't seem to realise that controlling the fundamentals of money (supply and price) causes wrong investment choices through inappropriate price signalling. And then they try and correct those choices by further fiddling with central control instead of letting the chips fall where they will fall. Because they're going to do this eventually - no amount of bailing out will fix a fundamentally out-of-date business model.

Lehman was indeed allowed to fail, and the crisis that its failure caused was sufficient to ensure that no government would be stupid enough to try that experiment another time. Of course, this is all theorising, but the general thinking in financial circles is that the government had to step in with monster bailouts to restore confidence precisely because Lehman was allowed to fail.

Wall Street organizations, in their present form, are functionally bankrupt. Having been bailed out, the clearing of bad debt never happened. This effectively trapped a large amount of capital in their vaults. In addition, the government is doing everything possible to funnel more productive resources into these black holes.

The real systemic risk was in throwing good resources after the bad. The "contagion" was other governments being spurred to do the same.

Some of this article is well written however the attacks on innovations such as ABS an CDOs is tacky, yes they haven't done well, but that's the nature of innovation. You're always going to have some innovations that do well and other that do badly, the only way you'll find out is through execution. They were both designed to solve practical problems in liquidity and price discovery.

Picking through individual failures and saying the whole system is bad on that basis is like saying startups are worthless because WebVan failed.

Hell of a place to choose to run 'experiments'.

Try that shit in an emerging market, first.

Each market is different. Even between developed markets, what might work in London might not work in New York or Tokyo.

Sure you can block all innovation, but you need to accept you'll block out the good as well as the bad. And that comes at the cost of economic growth.

It's a wildly unpopular idea here, but i think a progressive tax really encourages innovation.

A zillion small scale trades with little to no tax consequences encourage people really explore the space. Effective strategies grow till efficiency is overcome by tax friction. "To big to fail" is limited by the structure of the system, rather than regulation.

I wonder if it would be called innovation if we called these 'Products' what they really are. CDO='insurance on loans', Synthetic CDOS = 'making bets on other people's loans' When Las Vegas casinos come up with new ways to gamble is it called innovation?

Synthetic CDOs are actually more clever than that.

They solve the problem of me wanting mortgage bonds, and you wanting insurance on loans. We both want to take opposite positions (me long, you short). Without a synthetic CDO, we are unable to trade - I'll be trying to buy home loans from WaMu and you'll be talking to the AIG sales desk.

The problem with synthetic CDOs is that they're over the counter. You can have more people betting against loans than actual loans and since they weren't controlled the one who was supposed to pay out [1] might not be able to.

Speaking to the gambling point, I believe it was Deutsche bank who went to the government to make sure synthetic CDOs wouldn't be classified as gambling before starting to sell them.

[1] I can't say "issuer" because these were sold around.

> When Las Vegas casinos come up with new ways to gamble is it called innovation?

Yes.

In my personal opinion the difference between gambling and investment isn't in terminology. It's about expect return. If your expected return is negative than it's gambling, if your expected return is positive it's investment.

I really don't think gamblers think that way - I know people who play poker fairly seriously (which is clearly gambling) and they play to win and generally they do pretty well - certainly well above break-even.

I think serious poker is in a different category than slots or roulette.

You can call it gambling but there's a reason there exist top poker players who win with consistency. If it were really random chance, you wouldn't see players that make a living playing poker.

All gamblers play to win.

Think about it this way: if you buy a slot machine to put in your store/bar/etc, are you gambling or investing ? - it's clearly a game of chance, and you might lose money on it, however because the odds are set in your favour in the long term you'll make money. Hence it's an investment rather than gambling.

You do actually get angel investors who invest in poker players with solid track records.

Gambling is when the economic activity is zero-sum -- there is a loser exactly equal to the winner. Investment puts money to work so the effect is their is more real wealth in the world. For example, a farmer buys an irrigation system with a small business loan -- now there is more food in the world. A homeowner uses a home equity loan to insulate his house -- now there is more energy in the world.

Contrast this with a casino developing super-slots with a $1 million payout. The casino might increase their take, but their is no increase of real wealth in the world.

Well, if I go out and buy some stock in a publicly traded company post IPO the company won't see any of my money it will simply go to someone who purchase it from its current owner.

If there was no secondary market, the company would receive far less money at the IPO.

Liquidity is very valuable. Compare the more and less liquid shares in Chipotle - the only difference between CMG and CMG.B is that CMG.B has more voting rights, but less liquidity.

http://www.google.com/finance?chdnp=1&chdd=1&chds=1&...

Sorry - I was really thinking about the definition of "gambling" and "investing" above. So perhaps using that terminology the original investors at an IPO are willing to do so expecting liquidity based on the willingness of others to gamble on the stock.

Or something like that :-).

I agree, if you "invest" for a few months or days you are just gambling.

Check out this post on where to put your money now -- this is not a scam therefore it is not easy, or risk free:

http://scottlocklin.wordpress.com/2010/11/20/investments-for...

You're confining gambling only to casino games or games with vigs (http://en.wikipedia.org/wiki/Vigorish). Under your definition, when I play cards with my friends for money, I'm not gambling because my expected return is zero. Actually, judging by my known skills in any particular game vs. my partner's skills, what I'd be doing would be varying nightly.

What you're saying seems to be a useful distinction, though, but I'd say it's the difference between "risk entertainment" or something else and investment, rather than gambling and investment, because it's all gambling.

Unfortunately, the individual failures were bad enough to bring down the whole system.

tbh. If it wasn't CMOs it would have been something else.

The fundamental problem here is systematic risk, a bank should be able to fail without bring down other banks, unfortunately the system wasn't able to cope with the failure of a large number of banks who were all exposed to the same risk.

If a bunch of the big US car companies failed we may well have seen the same thing, because a lot of the US banks are heavily exposed to that sector. It just so happened it was mortgages that happened to be the tipping point this time.

Imagine your a bank and you have a bunch of other banks you deal with, you assign them all individually a "credit risk score" so you know how much extra to charge them to protect you against the risk of them failing before they have a chance to pay you what they owe you (much the same way as credit scores work for people).

But you have no-way of understanding the correlation between the risks, if all the banks your dealing with are actually exposed to the same underlying risk (i.e the CMO market) then the risk your exposed to is actually far higher than the sum of the individual risks. But obviously the banks you're dealing with can't tell you what their underlying risks are because that's propriety information, so all you have to go on is the risk rating given by ratings agencies.

This is a fundamentally hard problem to solve. No-one really has a good solution. It's much easier to say "let's ban CMOs" than to admit we don't really know how to solve the problem.

It would have been far cheaper just to rescue the FDIC and preserve the money held in savings/checking accounts.

Savings accounts are only a tiny part of the equation, the impact on businesses of a wide spread collapse of investment banks would be much larger.

Other countries which offered guarantees on savings accounts still bailed out banks due to the wider economic issues.

What is a scenario where business would be disrupted in this manner? After all, even with the bailouts, credit is as hard to obtain.

Company investment and banking accounts aren't protected by FDIC, so that alone could wipe out a lot of companies.

Credit would go from "hard" to "near-impossible". Investment banks finance all sorts of things you would never think about. For example the kitchen equipment in your local fast food chain (The pizza chain Dominos ended up buying the leasing arm of an investment bank which financed it's kitchens to ensure it's own stability).

Larger companies rely on their investment banks for all sorts of things, from FX to protecting against counter-party risk (i.e. providing insurance against your main customers or suppliers going bankrupt).

> The fundamental problem here is systematic risk, a bank should be able to fail without bring down other banks, unfortunately the system wasn't able to cope with the failure of a large number of banks who were all exposed to the same risk.

Note that some of the systemic risk was caused by regulation.

The US govt gave fannie and freddie stock special treatment when held by banks as part of their assets. As a result, banks overloaded on Fannie and Freddie. When Fannie and Freddie went down, that took a lot of banks into technical insolvency. (It didn't help that Fannie and Freddie lied about the loans in their portfolios, which threw off everyone's risk analysis of the market as a whole.)

Bond insurance was encouraged by regulators because it let banks and pension funds hold bonds as "safe" assets. (When you're pushing mortgages, you need to make them appear safe so more folks will buy them.)

The SEC gave a ratings monopoly to three ratings institutions. When they got it wrong....

Regulation is systemic risk.

Almost All of the systemic risk was caused by regulation. Too big to fail messed up with "the role of market discipline in financial markets" (3rd pillar of Basel II). Firms were even picking their own regulators (through loopholes) so they got the most lax ones. And so on.

I actually think that---since we cannot regulate for every contingency---then it's much better to avoid having too big to fail firms.

> then it's much better to avoid having too big to fail firms.

If a company that is, say 10x the size of Goldman Sachs, is "too big to fail" and therefore too big to allow to exist, what does that tell us about the US govt?

As I interpreted it, the attack was more on the economic rent extracted by the originators of the "zero sum" ABS and CDOs rather than the concepts themselves; the core thesis being that there would be a lot less profit to be extracted from originating opaque financial products if demand for them wasn't artificially inflated by buyers and sellers being insulated from the downside risk of their decisions.

It's more akin to saying you end up with too many startups getting too much funding in a climate where investors don't do adequate due diligence.

You won't find startups "innovating" in as brazenly cynical a manner as alleged in the Goldman Sachs example the author recounts either...startups don't profit from people shorting the stock they issue.

I think the point is that they actually don't solve any practical problems in liquidity or price discovery.

Does anyone on earth care about the price of lumber within even a 24 hour window? Why should we care about a sub second window? We don't. Wall Street refers to the high frequency game as "taking the dumb money". I call it "ripping off the value investors who actually serve a role in the economy".

Finance serves an important function in the economy, routing money to where it's needed. Are they doing a better job of that now than they were in 1985? Not really. So why are they extracting 10X the money in bonuses and profits?

I'm not sure about subsecond windows, but 24 hour windows certainly matter. Say you're planning on making a big order of lumber for your mill and you're ordering from a market with zero speculation so the prices jump up and down with the supplier's inventory. It makes a lot of sense then to wait a few days (during which your machines and workers are idle) to get the best price possible. I'm not going to say that high frequency trading serves a useful purpose, but to me that looks like a zero-sum struggle between Wall Street firms for profits that Wall Street deserves for doing its thing on a minute to minute or hour to hour level.

Ok, fine, downgrade 24 hours to 1 hour. Nobody cares about those prices within 1 hour. We've had markets that operate within 1 hour for over a century.

What did all the "innovation" actually produce? I know Wall St is making more money but are they providing more value?

High speed trading is only relevant to the competition between market makers and other short term speculators. It has no effect on long term speculators/hedgers/etc.

If you want to trade right now, you will pay a few cents/share for the privilege (this is the spread). If my company was fastest, we sit at the top of the order queue, and we will receive those few cents/share. If GS was faster, they get the pennies. Either way, you trade right now. And if I decide to cut in line by offering a better price, you trade right now for less.

Wall Street refers to the high frequency game as "taking the dumb money".

This is a rather strange thing for them to say, since it's usually the smart money types complaining about HFT.

It used to be that mutual/pension funds and other institutional investors could hide their big trades in noise - the market would adjust slowly, and other people (retail investors, smaller shops, assorted other suckers) would absorb the price impact of the trades. Now they are complaining they can't do this anymore.

http://www.advancedtrading.com/exchanges/showArticle.jhtml?a...

(The authors of this article work for a "rent our 'screw the retail guys' algorithm" outfit.)

Basically, assorted HFT shops catch them in the act, trade ahead of them, and split the profits with the former "suckers" [1].

[1] Crossing the spread means that the passive investors in the market are making a premium.

Well, you know a lot more about HFT than I do, I just know that I've heard several people refer to their trading practices as taking the dumb money, including a one-guy algo shop that sounded similar to how you've described your work.

At the end of the day though, I'm still stuck on my 1985 story -- compensation, profits, bonuses, etc have skyrocketed since then. But what good have all the newfangled instruments done for the rest of the economy? Doesn't seem like much. So is it just rent-seeking?

EDIT: RE: hiding the big trades in noise. What's wrong with that? Why should they have to develop a special HFT algorithm in order to simply move shares around without getting their lunch eaten? This is exactly the kind of value extraction I'm talking about. HFT shops have now created a huge expense for these retail shops that didn't exist before.. are the financial markets functioning better as a result? No, they're just sucking more money out as transaction costs.

Comp/bonuses went up for some. It went to zero for many others. In 1985, banks had armies of line workers processing trades, managed by officers making decent money. The average comp at a bank might be the average of 1 officer and 100-500 low level line workers. It would be better than other office jobs, but not great.

In 2010, the officer specs out business requirements after talking to traders and regulators, sends the high level work to IT and the CRUD apps to Bangalore. 475 of the 500 line workers have been replaced by a data center in Nutley, NJ, 5 remain in the US handling cash wires and other time-zone sensitive work, and 20 work via email out of an office park in Pune.

For obvious reasons, the average comp of 1 officer + 3 IT guys + 5 US line workers is going to be far higher than the average comp of 500 grunts + 1 officer.

Additionally, for many reasons (IT being one of the biggest ones), it's also vastly easier today to strike out on your own than ever before. In 1985, if you wanted to quit and start a hedge fund, it was tough. Today, a hedge fund can be operated by 3-4 guys + servers. Banks are forced to pay their traders enough money to prevent them from quitting and doing exactly this.

(Remind you of another high comp industry, with established players giving out big comp packages to retain talent?)

Edit (in response to your edit): RE: hiding the big trades in noise. What's wrong with that? Why should they have to develop a special HFT algorithm in order to simply move shares around without getting their lunch eaten? This is exactly the kind of value extraction I'm talking about. HFT shops have now created a huge expense for these retail shops that didn't exist before..

When you make a large trade, there will be a price impact. This is unavoidable - if you are selling millions of shares, supply went up, and price must drop.

In 1985, institutional investors (not retail investors) tried to make sure that other people (their counterparties) suffered the price impact. They would sell shares at full price, retail investors (i.e., my Mom) would buy from them, and then shares would go down once the market figured out supply dramatically increased.

In 2010, some HFT shops will sell to my Mom at a favorable price, and then buy from the institutional investor after prices go down. This benefits the HFT shop and retail investors at the expense of the institutional investor. I'm not saying this is good or bad, I'm saying it benefits "dumb money" (little guys who can't afford to rent an algorithm) at the expense of "smart money" (Goldman, Vanguard, Calpers Pension Fund).

[edit2: just a disclaimer - specific numbers/cities should not be taken as precise. The data center might be in Jersey City or Brooklyn, Pune might be Cebu or even Baltimore, and there might have been 250 or 1000 line workers rather than 500.]

This is an awesome convo BTW, thanks for providing all the detail about the industry.

Regarding our high comp industry vs theirs.. we've reinvented the world at least twice since 1985, and like I said, I haven't seen a reinvention of finance in any positive way. If they're that much more efficient now (and I agree that they are along the lines you're saying), where are the savings for the rest of the economy? I'm all about the TD Ameritrades of the world where I can do an $8 trade, that's providing value and they deserve every penny they earn. But it seems like most of the hedge funds are just playing games to extract money from less sophisticated investors.

Regarding your average comp comparison -- that's fine, I'm all about profiting from efficienbut what about the aggregate comp? Is that higher, too?

Regarding our high comp industry vs theirs.. we've reinvented the world at least twice since 1985, and like I said, I haven't seen a reinvention of finance in any positive way.

In almost every financial service that existed in 1985, margins have been cut to the bone (the main exceptions being M&A and IPO services). The commission on stock trades is now $0-8, it used to be $150, as you note.

When businesses need futures/options (mainly for hedging purposes), they often just buy them on the open market rather than contacting a GS/MS/JPM rep and paying through the nose for the privilege.

Regarding hedge funds/private equity, Warren Buffet and others have commented that there are very few hidden gems left, particularly gems detectable by financial/technical analysis. I.e., good businesses are getting more investment.

Regarding your average comp comparison -- that's fine, I'm all about profiting from efficienbut what about the aggregate comp? Is that higher, too?

It's a good question. Top traders get more, but they also tend to earn more. Good IT is certainly vastly cheaper than armies of line workers, and HFT is cheaper than human market makers. More financial services are provided today - my understanding is that credit cards were hardly pervasive in 1985, and no one would cell you a phone on quasi-credit like they do today.

I have absolutely no idea and I'd be very skeptical of anyone who claims to know.

Corporations can now protect themselves from all sorts of risks which weren't possible before using swaps and future instruments.

Innovations in debt financing drove the huge economic growth of the last two decades. Mobile phone technology would never have taken off if cellphone companies couldn't have used debt financing to build huge cell tower infrastructures.

CMOs for all the problems they caused, did allow for a huge increase in home ownership. For every default there are dozens of paid-up home owners who could never have afforded to buy a house otherwise.

Electronic Government bond auctions have driven down the cost of government borrowing, leading to lower taxes.

Forex costs have dropped massively. Look at your laptop. It's likely that during the manufacture dozens of countries were involved and likely hundreds of currency transactions. You paid less for your laptop because of a decade of innovation in the FX markets.

... At Goldman, it has been reported, nearly a thousand employees received bonuses of at least a million dollars in 2009. Not surprisingly, Wall Street has become the preferred destination for the bright young people who used to want to start up their own companies, work for NASA, or join the Peace Corps. At Harvard this spring, about a third of the seniors with secure jobs were heading to work in finance. Ben Friedman, a professor of economics at Harvard, recently wrote an article lamenting “the direction of such a large fraction of our most-skilled, best-educated, and most highly motivated young citizens to the financial sector.”

- The unintended consequences of the Federal Reserve system of monetary distortion.

It's gotten Michael Douglas at least 2 paying gigs.

Well, the Ferengi make regular pilgrimages to Earth's Wall Street, to them it's a holy site of commerce. That probably brings in a lot of tourism money.

Sources?

http://memory-alpha.org/wiki/Wall_Street

Didn't really read it all the way through, but for some reason, it is only me that finds it weird that the publish date is November 29, 2010? Is this an error or intentional?

Magazines (like the New Yorker) are usually dated a bit into the future, so that they look fresh on the news stands. At my local magazine shop, almost of the monthly magazines are already on their December issues, and some of them have been out for a week or two.

Have you really never noticed this before?

The New Yorker predates its issues, as they might theoretically take until Nov 29 to arrive in some people's houses (over here in brazil it takes longer than that, usually). It's an odd practice, but most magazines do it in some way.

What good is Google? or OkCupid? or any matching service?

For those who have nothing to seek, these services are indeed an environmental scourge. ;)

Reply on news.ycombinator.com