I agree the bulk of any volume is the market making, the hedging, etc. What I am referring to is what I saw on the team.... Let's say that you can make X returns for Y expenses for market making. Now imagine instead of market making you do structured products where you make X+1 returns with Y-1 expenses, where will you put your effort? Because I actually got to work hodge podge on many different types of projects and was lucky to talk to many other people I got to understand that most of the investment bank stuff is a FAD... You are just shifting money into different configurations. I think it was yesterday or today where a guy who started small but got to learn from floor traders and moved on learned that any automated strategy or technique or thing has a half-life. That means the way to make money is by constantly adapting and moving on. When I looked at the ib I was at, that was their attitude. Focus on the stuff that makes cash and if it stops making the cash move onto the next thing. Don't completely ignore it, but don't give them the money they think they need. Because after all if there is one thing that is for certain the moment a strategy starts loosing momentum what do we do? Oh we need to buy that tool, that utility, what have you. When the reality is that the profitability of the approach has died off and it is time to move on... This is why I saw high frequency trading in its current incarnation and its associated mathematics is a fad... And in about two years it will be time to move onto the next shiny thing... BTW what was the shiny thing three to four years ago at the ib I was at? Commodities... This year they basically shut down the entire commodities desk... Sure there are still some traders, but because the profitability went massively down it was time to move on... I have no idea what the shiny thing now is...
So if I understand well, you're essentially a programmer that built risk systems, as opposed to trading systems. Is that correct? You never traded or designed trading algos, never managed portfolios, never designed electronic exchanges, never studied in financial economics, or anything that could make you remotely understand HFT more than my mother does. Is that right? Arms races don't exist only when it comes to computers. All technology applied in such a way that it could make a business more simple to operate and eventually commoditize it is an arm. Wealth management is an arms race. Private portfolios are being increasingly built following model portfolios that are optimized and cloned. Institutional portfolios have been optimized for long, and risk models are increasingly similar from one firm to another. Stock selection models are increasingly automatized and proven to often render manual stock picking rather useful. And those models are also increasingly similar across firms. Fees are driven down, margins are tightened, people are getting fired, AUM is being consolidated, less people manage more money, etc. M&A is less and less of a contributor to the P&L of traditional investment banks. More firms are starting to rely on the services of smaller but equally prestigious boutiques that are less prone to conflicts of interest and information leaks. Of course, GS is still a leader, but their star bankers can always be inclined to leave to start their own firms, and clients could easily follow as it is a peoples' business - the firms commit no other asset to this business segment than their own employees. Everything else will eventually fall into the arms race category, including underwriting activity, which could easily be replaced by auction systems for new issues. Institutional investors will provide the bids and liquidity, and underwriters will go from risk takers to simple platform operators. And structuring deals, not an arms race? How far and deep can you go in thinking of new structures that fit a client's needs? How much value can useless complexity create? The recent crisis just revealed us how complexity can obfuscate things more than fullfill natural needs. Once you've thought of the few structures that naturally fulfill a client's financing, risk transfer, trading and M&A needs, and once the models, pricing formulae and hedging strategies have been written down, how can you say that deal structuring does not become an arms race? Truth is, the whole banking industry is technology driven and most of it is subject to an arms race one way or another. But hey, wait, such are most non-service industries also. How can you think that pumping oil from the ground is not an arms race? Or mining ore? Or producing clothes? Are these fads also? Wait buddy, I know the economics of an investment bank, thank you. Dark pools increase liquidity but do not improve price discovery at all, quite the opposite. They allow investors to trade with less market impact by avoiding to leak information about their trades. They are great for highly informed investors who want to build up large positions at the expense of less informed investors who mostly seek liquidity (ex. passive institutional investors). This means that uninformed, liquidity-driven traders have absolutely no benefit in trading in dark pools, as they are more at risk of trading against much more informed investors, at a very disavantageous price. On the other hand, trading in the open market allows them to execute at what the market believes is the fair price for a security given the uninformed nature of their trade. In the long run, this means that dark pools will attract only informed investors because of the opacity they provide, but will eventually fail to pool liquidity for the same reason - their lack of transparency. As for your 3% example, you sure must be kidding me. I don't know any portfolio manager who would be willing to pay a 3% trading cost on any trade. Not even in Ghana. Just think about what it would look like if index managers were paying 3% on their trades! What kind of fees would index funds carry? How much fees would you pay for a simple S&P500 replicating ETF? 5% per year? No, HFT works because there is natural demand for liquidity and price discovery. The fact that some are getting lots and lots of chips wheras others don't is just the nature of the industry. You eat what you kill, the winner takes it all, survival of the fittest, etc. You've heard all these before. This is corporate america. And no, it's not a fad.
Well, I guess they closed the commodities desk because they failed to capture the pole position, did they? Maybe they lost to someone else? Maybe they should ask Morgan Stanley how things are going on their energy desk... In two years from now, you'll see lots of people moving on because they will have failed to capture the pole position in HFT. But you probably wont see Citadel, Goldman, SAC and Renaissance on this list.
The only constant in this game is change. Two cash cows of the HFT game might become extinct, dark pools and FLASH orders. That's gonna make the market shake off a few players... Those who remain are going to be in need of original ideas... Then someone's gonna come up with a new way to find a real edge in the market (usually by throwing money at the problem till someone figures it out), it's the way it's always been. Just like Livermore said, "there's nothing new under the sun, especially not in Wall Street". The toys might change but the game remains the same.
Thank-you somebody who gets it... I especially liked the comment: Those who remain are going to be in need of original ideas... It is that step that costs lots and lots of money... Not that it cannot be done, but it costs. The IB I worked for did not want to be number 1. They prefer to make money because at the end of the day that is what it is about, making money. The CEO who ran the IB was a clever man who preferred the understatment. The motto was to strive for things like "best" than "number 1". The thing that the IB is trying very hard to get going is the ability to react. They want to be able to go from trader -> quant -> software in a very fast and flexible manner. It is not easy for them, but they want to be able to react very quickly to whatever comes up.
For the record... The IB I worked at did not receive government money! And their profitability has been pretty good considering how big the mess has been. So I guess they did something right...
Take it easy, Christian. That's the usual welcome scenario on ET. Some ET members have long standing experience in strengthening each others backbone... Your comments are very welcomed as they show that IB's are constantly rethinking their business models, too - and that they have to adapt, too. Like everybody else on this board, or ? May I add that the "arms race" in HFT is not only a technological issue but also a "fire power" issue in terms of available "risk capital". HFT is only practicable for IB desks and "hedge funds" ( it's highly questionable to name these players hedge funds anymore ) with above average risk capital exposures. Here is some comment from Tabb group regarding the alleged stolen code by Sergey Aleynikov, but if you read between the lines alot of questions arise : http://advancedtrading.com/algorithms/showArticle.jhtml?articleID=218401501#undefined Have fun on ET !
Two comments. One, I disagree HFT is only practicable for IB desks and "hedge funds", sure it takes a certain lvl of capital investment (as you say, "Fire power") and fixed cost platform building to play in the HFT game, but it is not all that high, really. We are not talking about 100M or more, there are quite a few firms who are quite successful in HFT with 1-10M of capital. Let's face it, if you are doing 1000s of trades a day, the amt of capital you can allocate to each trade can not be gigantic by definition. As a tangent, I do believe the media is overblowing how profitable HFT can be, and it is not just a matter of "us" (for disclosure, I run a HFT outfit) not wanting other ppl to crowd our space, it is just that the amount of capital we can deploy is really very limited. Which, of cos, makes what misha did all that more impressive. So since the capital is limited *and* the startup fixed cost is not all that high, I truly believe HFT is something that smaller firms can compete in (I certainly believe so! since this is exactly what I do on a daily basis). And frankly, what my firm is doing, is not all that "special", I have joked that our "secret sauce" can fit on a single page recipe card, and still have room left over for the baked brownie recipe. It is just how the HFT systems "model" the market, that's all. I am going to use an overused example here, RenTech Medallion, which, by some estimates, is only about 6-7B in size. Why won't he open up Medallion to other investors, and charge a SAC like 4/50 fee? It is because the strategies have size limitations. And from RenTech launching the RIEF and RIFF, it is not that Simons thinks "he made enough", he is definitely greedy and wants more. Second comment I will reply to nitro's comment.