I've just read Krugman's recent op-ed on high frequency trading, and, as much as it pains me to say it (I am a fairly big Krugman fan), Krugman is wrong.
Updated at the end.
First of all, there is a difference between high speed trading and "front running" or "forward trading". Front running is when your broker uses their position of trust (by you) to trade against you. In this case, your broker has special information not available to the public or anyone else- information about your trades before they are submitted to the exchange. And guess what? Front running is already illegal! So what we need is not new regulations, but rather to enforce the regulations we already have. If you want to crack down on front running, I am with you. And so to, I note, is most of Wall Street. Most of Wall Street considers front running (which hurts their profits!) to be incredible scummy.
But not all high speed trading is front running- in fact, hardly any of it is. And this is where Krugman's article breaks down. All high speed trading means is that the people/companies doing the trading are spending the money, time, and effort to minimize the delays on trading.
For example, the average person has this thing called "a day job", which generally doesn't involve sitting around and trading stocks all day. The big trading companies hire hundreds, nay, thousands of people whose job is to do exactly that. Who has the advantage- the person who has something else they're supposed to be doing, or the person who's job it is to trade stocks? So is Krugman proposing that everyone should become full time stock traders, to overcome the disadvantage of those having to work for a living?
Or, consider this. The speed of light is 186,282 miles per second, and it is 1,018 air miles from New York to Minneapolis. This means that it takes almost 5.5 milliseconds for light to travel from New York to Minneapolis, one way. That means it takes at least 11 milliseconds for a packet of information to travel from New York to Minneapolis, and the response to travel back again, assuming no delay in Minneapolis. A computer sitting in Minneapolis is going to be at least 11 milliseconds slower than the same computer sitting in New York, for trading on a New York exchange. And 11 milliseconds is a long time for a computer these days. Is Krugman proposing that we abolish the speed of light?
High speed trading is not a cheap market to get into, either. Even ignoring the people costs- the programmers to programming the trading software, the admins to administer the computers, the quants to come up with the trading strategies, the traders to monitor the trades and make sure they're not going bonkers, and the support staff for all of the above- it's going to cost hundreds of thousands of dollars to get high-speed, low-latency connections to an exchange. In addition to the telecom company's charges (we're not talking a DSL drop here, remember- we're talking dedicated hard lines), you have the charges by the exchange itself (which aren't all profits- every high speed line they have to deal with means more infrastructure they need).
Krugman utterly fails to show a case as to why high speed trading- not trading based upon taking advantage of positions of trust and trading on information that really isn't public domain, but traders who simply spend the money to be in an advantageous position relative to the market by virtual of physical locale (being in New York and not Minneapolis), investment in infrastructure, or the ability to dedicate time (this is their day job)- harm the market.
What high speed traders do is make the market more efficient. An assumption a lot of economists make is that markets are perfectly efficient. This means that if a change happens, the market will "instantly" reflect that change- otherwise there is an arbitrage trade (a "no risk" trade) that can be made against the market. Now, no market really is perfectly efficient (in a perfectly efficient market there is no reason to trade)- but high speed traders do tend to make the markets more efficient, by increasing the speed in which they respond to changes, and keeping the "error", the difference between where the market is and where a perfectly efficient market would be, small. The average person doesn't get to take advantage of the small arbitrages the high speed traders take advantage of- but they also have a guarantee that there isn't a large arbitrage that can be made against them. The average trader may lose a few pennies per share vr.s a theoretical perfectly efficient market, but the loss will be kept to a minimum.
The con that was pulled on the American people by Wall Street wasn't that we allowed high speed trading to exist. No, the con was that we were told we had a prayer competing against the big boys in terms of sheer speed. You can't. Any more than you could start your own car company. Not because some evil conspiracy (by the government, the rich, the illuminati, whomever) is preventing you, but simply because the barrier to entry is too high. If you're trying to do day trading and you don't have the money to do it as your day job, buy bloomberg terminals for everyone, rent an office in the financial district of New York, etc., you are going to lose to the people who can afford to do all those things. You're bringing a knife to a gun fight in this case.
This isn't to say that you can't make money trading the stock market, you can. I'd like to point out Better Investing as a good resource for people (full disclosure: I'm a member of a BI, and my mother is on the national board). The Better Investing strategy presumes that you can't compete with the big boys on speed, and thus shouldn't even try. It's a long-term (3-5 years) hold strategy- you buy the stock, sell it five years later for twice what you bought it for, and make a nice 15%/annum return. I'm simplifying a fair bit here, and leaving out the secret sauce of how to pick a stock that is likely to double in five years- go check out BI for that.
If you're buying the stock for $15/share, and selling it for $30/share, the fact that you lost a few pennies per share buying the stock, and lost a few more pennies selling the stock, because you haven't spent the money to be a high-speed trader, doesn't matter. So you don't need to be checking the stock every couple of second- if fact, you shouldn't be doing that. All you need to do is check in with the stock once a month, or once a quarter. Rather than being a full time job, it's a evening or two a month hobby.
I am a huge fan of Paul Krugman. I think "The Great Unraveling" and "The Return of Depression Economics" are required reading for everyone. I generally look forward to his columns. But this column is a huge failure on the part of Krugman- he is just wrong. Now I have a job to get back to.
UPDATE
Well, the comments are comming in, and there are obviously some trends developing, so I thought I'd address them in bulk.
First, obviously I haven't denounced Trotsky, er, Wall Street, enough. I missed the rule that every diary that mentions Wall Street has to have, as it's central premise, the Wall Street bonuses. I actually do believe that if your company is receiving, or has recently received, government bail out money, you don't get to pay pay bonuses (everyone's bonus is that they get to keep their jobs). But I mistakenly wrote a diary about high speed trading, and not the bonuses. My bad.
Second, it is very clear that large numbers of you don't realize why an efficient market, even an approximation to an efficient market, is worthwhile to everyone. Basically, the definition of an inefficient market is that people are getting screwed- people are either buying things at too high of a price, or selling things at too low of a price. Now, the "problem" is that you, the low speed trader (and yes, if you take a full second to process an order, you're a low speed trader) don't get in on the screwing. By the time you can respond to a market inefficiency (someone wanting to buy too high or sell too low), the high speed traders have already moved in and completed the trade.
But the advantage is that the existence of the high speed traders, and especially the competition between the high speed traders, keeps the sizes of these "screwings" small. For example, let's say the "true price" of a stock is $35/share. Which means that theoretically you should be able to both buy it and sell it at $35/share- IF the market was perfectly efficient. But the market isn't- which means to buy the stock you'll need to pay more than $35/share, and to sell the stock you'll get less than $35/share. The question is how much that difference is. Would you rather buy the stock at $35.10/share, or $35.01/share?
The problem is that the "true price" of the share is not a static value- it changes second to second, due to the normal volatility of the market. It doesn't change that much- it's highly unlikely that a stock woth $35 now will be worth $25 bucks a second from now (not unheard of, but not likely). On the other hand, the stock moving from $35 to $35.10 in the course of a second or two is pretty common. This is where the high speed becomes important. Let's say the rules are changed so I can't change an offer (to sell a stock at a given price) more than once a second. If there is a good chance that in the next second the stock will be worth $35.10, not $35, I don't want to put out an offer of $35.02, or even $35.08, because if the stock does move, I get screwed over (I just sold a stock worth $35.10 for $35.02- opps). So instead I put out an offer of $35.12- and the person who wants to buy the stock is now paying $35.12 instead of $35.02. Being able to get into- and out of- the market fast (high speed trading) allows me to trade closer to the "true price" of the stock. I'm still making an arbitrage profit- but a much smaller one. If a high speed trader were to try and take a bigger advantage (say, selling the stock at $35.03 instead of $35.02), some other high speed trader would jump in and sell for $35.02.
The same logic goes for a tax per trade concept. If there is, say, a $0.05/share tax on trade (which was proposed in the comments), then I can't offer to sell a $35/share stock at $35.02, I've got to price it at $35.07, just to cover the tax. All the tax has done is widen the spreads- and screw over the low speed trader yet again.
Third, a lot of people are mad that the high speed traders have "bought themselves an advantage". Grow up people- this is true of every company that takes advantage of mass production or economies of scale. Ford Motor company can- with their bought advantages of big, efficient factories, make cars cheaper than you or I can make in our garages. Big efficient factories should be outlawed! Um, no. In fact, the barrier to entry to become a high speed trader is lower than the barrier to entry to become a car company! You can start a high speed trading firm with only a few hundred thousand dollars of capital expenditures and a few million of seed money- try starting a car company for that little.
There are a lot of things that do need to get fixed with our financial markets. And Paul Krugman is, IMHO, the go-to guy for what needs to get fixed and how. But in this specific instance he's wrong- high speed trading does produce a value to society, and shouldn't be banned.