NEW YORK (Reuters Blogs) -- Stevie Cohen is one of the greatest stock-market traders of all time. Indeed, there's a strong case to be made that he's the greatest. Cohen is not the greatest investor -- he doesn't really go in for buy-and-hold positions which steadily accumulate enormous value over decades. He's not even the greatest hedge-fund manager: he doesn't go in for the big macro bets (Soros vs. the pound, Paulson vs. mortgage-backed securities), which are the stuff of legend. Instead, he's a trader, and while normal people pretty much understand what someone like Warren Buffett does, or what someone like John Paulson does, it's much harder to understand what a trader does, or what differentiates a good trader from a bad trader.Trading isn't usually about making bets and then cashing them in when things go as you thought they would. It's more about understanding probabilities, seeing when securities are mispriced, taking advantage of fleeting arbitrage opportunities, being paid for providing liquidity to the markets (selling when others are buying, buying when others are selling), and, most importantly, "reading the tape" -- understanding the way that money is flowing around the market, and how those flows are going to manifest themselves in securities prices. Being a great trader is hard work: You've got to be constantly aware of subtle price actions in dozens of different markets and thousands of different securities. ( TheStreet's Jim Cramer is a great example of a trader: He doesn't have a deep understanding of any particular stock, but he knows where thousands of them are trading, and how their movements relate to each other.) What's more, trading is hard to scale effectively. You need to be a certain minimum size in order to be effective, but there's a maximum size too: You have to be able to get in and out of positions without moving the market so much in doing so that you end up erasing all of your profits. Being a great trader is also increasingly difficult. Thirty years ago, for instance, you could make surprisingly good money with very, very basic strategies. You could buy convertible bonds at issue, for instance, and hedge by shorting the underlying stock; or, even more simply, you could just pick a set of stocks and buy consistently at the bid while selling consistently at the ask. The buyers and sellers would pretty much cancel each other out, and you'd pocket the bid-ask spread, which, in the years before decimalization, was often substantial. Today, however, all of those strategies have been arbitraged away by algorithms, and the result is that markets are faster and more treacherous than ever. Strategies which seem as though they're work very well often have enormous and unforeseeable fat tails: Look at the monster losses during the quant meltdown of 2007, for instance, or JP Morgan's crazy London Whale trade. And yet there's still one thing which can scale, and which will never be competed away by algorithms, and where the upside is much larger than the downside: black edge. Cohen has never been easy to invest with. He deliberately charges some of the highest fees in the industry -- his 3-and-50 makes the standard 2-and-20 seem downright generous. And even then it has historically been very hard to get him to agree to manage your money. Cohen makes his fund inaccessible for a reason: He knows how hard it is to scale the astonishing results he's been posting, year after year, and that at the margin, the bigger he gets, the lower the returns he's likely to see. But at the same time, there's no way that he can run a $15 billion trading book on his own. He has roughly 1,000 employees, of which about 300 are investment professionals. And if you're one of those professionals, you have one of the hardest jobs in the business. The way that SAC works is that Cohen gives his individual traders, and teams, their own trading accounts, with millions or billions of dollars. The traders who make the most money get the biggest allocations. Traders get paid a percentage of the profits they make, which makes them compete against each other: in order to be successful at SAC it isn't good enough to make good profits. Instead, you have to make better profits than any of the other traders -- who themselves are some of the best in the business. If you can't do that, you get fired. If you can do that, you get to manage ever-increasing amounts of money -- plus, Cohen will mirror your positions in his own account, the largest at the firm, giving you a shot at extra profits over and above the ones generated by your own positions. In the immortal words of David Mamet, first prize is a Cadillac El Dorado. Second prize is a set of steak knives. Third prize is you're fired. While Cohen does still generate his own ideas, then, most of the time he outsources that function to his employees. There's a relatively static allocation of capital between the various traders, but then there's a dynamic overlay as well: Cohen "tags" the positions in his own account with the names of the traders whose trades they are, thereby giving every trader the opportunity to see his positions multiplied in size at any time. While his traders are moving money in and out of stocks, Cohen can be thought of moving his money in and out of his own traders' positions. He's not betting on stocks so much as he's betting on individual employees, in one big zero-sum game. As such, Cohen is much more than a simple employer/supervisor. He's constantly sending clear and public messages to his traders, about what he likes, what he approves of, and what he disapproves of -- and he's sending those messages in the most unambiguous way possible, in the form of extraordinarily large sums of money. If he wanted to, he could withhold money, and even employment, from anybody who was working with black edge. Alternatively, he could manufacture a spurious layer of deniability, while actively encouraging, in terms of financial incentives, the one kind of trade which has the very best risk-adjusted returns. The SEC's decision to charge Cohen with failing to supervise his employees is, yes, a cleverway to try to put together the most winnable case before the statute of limitations runs out. But it's also a serious charge which goes straight to the main way in which Cohen makes his money. Cohen's returns come directly from the way that he supervises and incentivizes his employees, and once you've read the complaint, it's pretty clear that Cohen loves any trade which makes money, and has no particular compunctions when it comes to whether or not the trader in question is behaving in an entirely legal manner. It'll be interesting to see Cohen's defense to these charges, but he has an uphill task ahead of him -- especially given that the hearing will be held in front of the SEC's own judge. The SEC has home-field advantage, here, while Cohen oversees a firm which has seen four different traders already plead guilty to criminal insider-trading charges. It's good that the SEC has finally managed to charge Cohen personally, rather than just his traders. The only pity is that we're still a long way from a criminal case. That might yet come, but I'm not holding my breath. -- Written by Felix Salmon in New York. Read more of Felix's blogs at Reuters.