CHAPTER 9

Trading with a Real Edge

The concept of an edge in the financial markets is as widespread among struggling traders as it is among full-time prop traders. The difference I have learned over the years is the interpretation of what makes a real edge.

A segment of struggling traders are holy grail chasers. If you've come across any trading book in the past 10 years, you've probably heard the old statement “there is no holy grail” a few million times. And whether it's openly admitted or not, most struggling traders continue to search for that one, single, elusive holy grail strategy.

Here's the real secret: There's no single strategy, no single edge that can be exploited forever. There are, however, many types of edges that come and go, and that's more than enough for a full-time trader to make a living.

Turn a different perspective onto this. If General Electric taught you “the secret strategy to becoming a global conglomerate,” do you really believe every entrepreneur should spend his whole life trying to replicate the exact process that GE used to become what it is today? No. New companies like Google have had no trouble becoming massive global enterprises by exploiting their own edge in the market.

The point isn't to stop looking for a real edge in the markets. The point is that your job as a trader should be to find and develop an edge that works today, in the market that you will be trading for the coming month or year. And when that begins to weaken, you adapt and develop a new one. Remember, half of the businesses that GE owns today are using social networking marketing techniques that didn't even exist more than a decade ago, let alone when the parent company was founded.

Likewise, don't label yourself a momentum trader or scalper and stubbornly spend every single day repeating the same thing with diminishing returns. Continue with your core strategies if they still make part of your income and you're comfortable with them, but always keep your eyes open for new ideas that might give you 90 percent of your paychecks for the next two years.

One of the most important concepts I want to convey with this book is the difference between a hard edge and a soft edge.

By my definition, a soft edge is the typical goal of most traders. Soft edges are all that most traders will find in the majority of their careers in the markets. And there's nothing wrong with that. The core strategies of most full-time traders are built on soft edges. Almost all the core strategies that have been employed by the majority of hedge funds actually fall into this category, as well.

A soft edge is one that's good enough for a trader to make money with enough discipline, consistency, and skill. Since the real mathematical advantage behind such an edge is rarely more than 1 to 5 percent above a raw 50-50 bet, a very large dose of discipline, consistency, and skill should be developed to properly exploit these soft edges.

At the risk of stepping into stereotypical trading guru-speak territory, it's generally true that enough hard work and discipline can make a trader profitable. What most won't tell you is that the degree of necessity of these traits is greater when your strategy's edge is just a soft edge.

A hard edge, on the other hand, is an inefficiency discovered in the market, at any given time, that can be exploited by anyone from a trained monkey on a keyboard to a fresh college grad who spent more time studying alcohol than algorithms. Hard edges, of course, are far rarer in any market—but when they fall into your lap, for crying out loud, don't miss the opportunity to exploit them as much as possible while you still can.

Keep in mind that the specific type of hard edges I'm discussing here would actually not be illegal but are often on the very edge of the cliff of compliance. In other words, they might be outlawed or circumvented within a few years. And some traders might even argue that the reasons given for outlawing such things actually hold less merit than the justifications given for many commonly accepted practices by the largest firms in this industry, but remember: Debating politics and regulations is for politicians; your job is to find a new edge if your last one is dying out.

Just don't rest your entire career and future on a single hard edge. Ethical or not, the opportunity to exploit most real edges will come and go regardless of the nature of it. Learn to be comfortable with diversifying your strategies instead of putting all your strategic eggs in one basket, so to speak.

I've come across a few people in this business who insist on treating this line of work like a hard science. These are the types of people who insist that no strategy is worthwhile unless it's been back-tested on upward of 10 to 20 years of data. The concept behind such a statement is all well and good, but the truth is that a strategy tested on historical data of entire decades is still only worth as much as the remaining time in the future that the strategy will remain effective to the same degree. In other words, you can test a winning strategy on every tick of historical data from 1970 to 2011 and show extremely an small average loss to average profit ratio, but there's still not one ounce of a guarantee that the strategy won't produce a much larger average loss to average profit ratio tomorrow or next month.

Having spent most of my career as a day trader, I've also learned that there's often nothing wrong with taking short-term profits by exploiting what I like to call transient edges. Some of these may be patterns or specific, predictable statistical likelihoods that have been found to be exploitable by a strategy within the past year or sometimes even less. For all intents and purposes, such a criterion would set a trader up to trade a strategy that many would consider not nearly robust enough to consider in the long term. And that's fine. While they spend the next two months looking for patterns in decades of tick data, you and I can actually make money on a day-to-day basis using something that happens to work right now—knowing full well that it may stop working at any time. As long as such a strategy is not overemphasized in your mind to the point where you consider it your staple trading strategy, there's really nothing wrong with collecting short term paychecks for something that may not last much longer than a few months.

Money is money. We're here to trade for profits, not to search for self-indulging scientific and statistical discoveries for the next Long Term Capital Management (LTCM, the hedge fund whose losses nearly brought down the global financial system in 1998).

Don't get me wrong. There's nothing entirely wrong with searching through ages of tick data to discover a statistical relationship that might help you build the basis of a very robust trading strategy. My point is that sometimes the small transient statistical patterns are just as worthy of a day trader's attention (as long as the trader always remembers what type of strategy he or she is actually working with in these cases). On top of that, it's also worthy to note that even if and when you find an extremely robust (or apparently robust) statistical relationship or pattern that worked over the past four decades, you should still treat it with at least some of the same caution that you would for a transient pattern you found to exist only in the past two months of data. Why? Well, there may be less of a chance that such a pattern might stop working tomorrow or next month, but it's still not impossible, as many an overconfident hedge fund manager has discovered over the years.

All too often, people are cautious about the wrong dangers and overconfident in the wrong situations. Statistically, it's far more likely for a person to end up in a car accident than in a plane crash, yet boarding an aircraft tends to cause far more anxiety in many more people than the act of getting into a car's driver seat. No matter what sort of edge you've found to exploit with your trading strategy, don't be too afraid to pull the trigger just because of past evidence of apparently robust and time-testing strategies that abruptly turned around. At the same time, don't be so trigger-happy that you start ramping up risk to unreasonable levels for your strategy.

Most of all, be honest with yourself. Are you trading on a hard edge or just a soft edge? You'll probably get to your destination eventually either way, but don't speed like a maniac on icy winter roads and, equally, don't be afraid to speed up a little on those warm, ice-free days when you know you've got excellent traction on the highway.

* * *

Now for the basic method of scalping. One of the regulations in place on the U.S. equity markets in recent years is that for stocks presently priced above $1.00, the minimum distance between the displayed national best bid and national best offer was one penny. Meaning, if a stock is currently trading around 12.10 and the highest displayed bid on any of the visible venues (NYSE floor or ECN systems such as NASDAQ, NYSE Arca, BATS, Direct Edge, etc.) is posted visibly at 12.10, then the lowest displayed offer that can be posted is 12.11 or higher. In other words, the spread can widen above one penny but it can never be tighter than one penny. Now, you might have noticed that I've purposely repeated variations of the terms visibly and displayed throughout the description of this rule—and that's a key to a concept used by many of the high volume scalpers in recent years.

It's against regulations for the displayed NBBO to be tighter than one penny, but not for hidden non-displayed or dark liquidity. The reasons and merits of this are beyond the discussions of this book. The important fact here is that trading the midpoint of the NBBO has become a major asset for those who can maneuver with it and can, by itself, be something of a soft edge in the markets or even contribute as a component of a hard edge.

Let's consider first that the main challenge of a scalper is the ability to get on the train when the train is headed in the direction you want to go. If this concept escapes you, I'll review the basic logic of scalping at the full penny prices on a highly liquid stock such as the names among the top average daily volume lists on the NYSE and NASDAQ websites. (The symbols in these lists regularly change due to a number of factors but it should be relatively simple to find the most up-to-date list at any time.)

We'll start with an example scenario. A stock is currently trading with a national best bid of 12.01 and a national best offer of 12.02 (in other words, the quote is “12.01 by 12.02”). On the Level 2 quote window, you see:

12.01 ARCA 250,000	220,000 NSDQ 12.02
12.01 NSDQ 240,000	210,000 ARCA 12.02
12.01 BATS 235,000	245,000 BATS 12.02
12.01 EDGX 120,000	110,000 EDGX 12.02
12.01 EDGA 100,000	100,000 EDGA 12.02

In this Level 2 quote, the displayed venues (electronic exchanges and ECN systems) are sorted in the order of the largest number of shares posted on them at a particular price, and are displayed with the national best bid on the left and the national best offer on the right. In real time, they would change every time a new trade takes place as liquidity removers would effectively be taking these shares off of the displayed orders. For instance, if a trader were to enter a market buy order or a limit buy order for 1,000 shares with the price 12.02 to NYSE Arca (ARCA), then the tape would print a 1,000 share transaction on Arca (P on the tape) at 12.02 (which, as long as it remains the national best offer price, would likely print in green on most platforms). As a result, the displayed shares on Arca would reduce from 210,000 to 209,000 as 1,000 shares have been removed from it. In reality, new orders will be posted at different times while trades like this take place so this event may occur with a less obvious example, but this is typically how liquidity removers interact with displayed liquidity adders.

Now, for a traditional full-penny scalper, what is most important to look for is when the numbers begin to reduce on one side but not the other. At any time, you'll see the number of shares reduce then increase; increase then reduce; pause a little and repeat. Much of this action may be aggregated orders from retail brokers who don't offer direct market access, orders from large institutions (including hedge funds) who may have a very different long-term time horizon, as well as fellow scalpers with the same time horizon. In all cases, this interaction will show up the same way, and much of it will be dismissible to some extent. What is significant is when you see large order after large order following each other like a wave, and whether they originate from human traders or algorithms programmed by humans should make no difference to a real trader of either kind—these waves will signal a micro-term trend that can never be seen on a chart. as it never actually moves the price up or down.

It shows what is about to happen before any chart in the world can register that anything has happened at all.

In this example, say there is nothing but semi-random action on this stock for the last half minute. The bid remains at 12.01 and the offer remains at 12.02. A few trades take place at 12.01, which prints red. A few happen at 12.02, which prints green. Red, red, green, red, green, green, red, green. There's no rhyme or reason, and no real pressure on either side. All you see is what appears to be pure randomness with orders that may have any time horizon known to the markets.

Suddenly, a lot of red prints hit the tape. 10,000 shares hit the bid at 12.01 and then 5,000 shares hit the bid. And so on. Rather than a Christmas tree (in the case of this particular trading platform) color scheme, it's turning into a red blood bath while the number of shares on the bid at each of the displayed venues begins to shrink. This is key—this scenario is only significant if the tape is combined with the sizes on the Level 2 quotes. (Likewise, if this were a scenario where the tape was printing green over and over, and the sizes on the displayed offer instead of the bid were shrinking, then the greater likelihood is that the market will move up to the next level in the next half minute.)

Traditionally, one might say that a series of sell orders triggered the psychological reaction of all the other market participants to jump on board for the move down to the next level (in this case, 12.00 by 12.01, which is one penny down from 12.01 by 12.02, where we began this example.) Today, it might be more realistic to say that some criteria, including an initial series of sell orders, which may have originated from humans at a keyboard or automated algorithms executing orders that may have any time horizon (or a combination of both), triggered other automated algorithms as well as human traders to jump on board and attempt to follow the stock down to the next level. In either case, whether you're a human trader or an automated trading algorithm programmed by a human who understands this aspect of this market, the obvious course of action is to bet on the stock moving down to the next price level—12.00 by 12.01 in this example.

This is a basic dynamic of the micro-term market action that cannot be viewed on a traditional price chart: The scenario is bearish when trades (prints on the consolidated tape) keep hitting the bid price (typically red prints) and the number of displayed shares on the bid is shrinking. It looks bullish when trades keep hitting the offer price (green prints) and the number of displayed shares on the offer is shrinking. However, it's often difficult to secure profits using this information alone, as every other human trader (as well as every other active algorithm) sees the same information that you see.

In the scenario just described, the action looks bearish (red prints, shrinking displayed bid) but would it make sense to aggressively jump on board by selling to the bid price, too? Maybe. But if you do so, remember that you would be removing liquidity by definition—you are selling (or short-selling) to the bid, and therefore you will likely be charged by the ECN system or electronic exchange for doing so. (Likewise, if it had been a series of green prints together with a shrinking offer instead of red prints with a shrinking bid, then you would be doing the same thing by buying from the offer price—still removing liquidity, by definition.) Meanwhile, in most cases, the party who posted the order that is being matched against yours, in this case the limit buy order that is posted on the bid price to one of the ECN systems or exchanges, will receive a rebate for having added liquidity to the market (the liquidity you just removed).

So even if the market does, indeed, move down to the next level at 12.00 by 12.01, you will then have to watch for two scenarios: Either the market turns around and returns to where we were before (12.01 by 12.02) or it continues falling (11.99 by 12.00), or it remains here (at 12.00 by 12.01) for an extended period before either of those scenarios happen. However, a trader is not a psychic or an idiot—your job is not to guess, but to watch the actual live view of the market in micro term: the level 2 quotes and the Time & Sales consolidated tape prints. All the possible scenarios at this point would be fine even if the odds were completely even between up and down, but there is the issue of your liquidity interaction: Will you be able to enter and exit without being charged an ECN fee on both sides or even one side?

If you question why this matters at all, consider this: If the market continues to head down, you can post a buy order at 12.00 to exit for a one penny profit. (At 10,000 shares, this is a $100 USD profit.) If the market looks to be reversing and headed back up, you can hit the offer before it breaks and get out at breakeven. If you trade once or twice a day, this may not matter, but no strategy that relies on this small a price movement can be traded consistently with only one or two trades per day. Instead, if you really scalp this way, you will find that your ECN fees (for removing liquidity) will quickly add up. For example, if you entered the trade by hitting the bid (sold short to the bid at 12.01) on Arca and then exited the trade at breakeven on NASDAQ when it looked to be reversing back up (bought to cover from the offer at 12.01 when 12.01 is the new offer), you will pay the fee to remove on Arca this month and also the fee to remove on NASDAQ this month. Typically, the fee on both sides is close to $3.00 per 1,000 shares (that's a $60 loss on a 10,000 share trade that aimed to make $100 per profitable trade).

Ideally, a trader would be able to get filled passively with an add liquidity order and get filled right before a favorable movement. For instance, if the market is currently at 12.01 by 12.02 and it begins to print red, the ideal entry would be to see it print red and see the displayed bid sizes shrinking, and then enter a passive sell limit order at 12.02 to short sell from that price. As the market follows through and breaks down to 12.00 by 12.01, your short sell entry at 12.02 is now ideal. Likewise, if the market was at 12.01 by 12.02 and you saw a series of green prints at the same time the displayed offers on the Level 2 began to shrink, then ideally you could post a limit buy order to the bid at 12.01 and get filled with a long position from 12.01 as the market breaks up to 12.02 by 12.03. However, such a scenario is, in most realistic cases, akin to wishing for the ability to bet on a winning team within 10 seconds of the end of a ballgame. It would be ideal but realistically close to impossible in most cases as no other participant (human or automated) will likely be stupid enough to buy from your offer when the market is printing all red and the bid is shrinking; likewise, no one is stupid enough to sell to your bid when the market is printing all green and the offer is shrinking. Even if a few parties were, theoretically, that stupid, keep in mind that most ECN systems work based on a price-time priority. In other words, when the market is printing green and the offer is shrinking, there is no way the bid orders that were there before you would be shrinking (the humans and automated systems who posted those bids before would keep them posted when it looks this obvious that the market is about to head up), so if you add your own bid to try to get in on the train upward, you'll be way at the back of the line. A couple of idiots who might hit the bid price still would not fill your order; it would fill the orders of the bidders at the front of the line. (And, of course, likewise for a short sell order at the back of the line of the offer side when the market is printing red and the displayed bids are shrinking.)

Instead, a scalper must balance predictability with the ability to get his or her order filled, and also manage ECN fees or outright aim for ECN rebates.

Scalpers learned long ago that rebates would be a major asset to a trader. For instance, if you add liquidity on any of the major displayed venues (like Arca, NASDAQ, or BATS), the ECN would typically pay you a rebate of something in the neighborhood of $2.00 to $2.50 per thousand shares. ($40 to $50 per 10,000 shares round-turn in a trade that aims to make $100 per penny at this share size.) The art of trading for rebates had long become one of the traditional staples of high volume scalping at Swift Trade, but the dynamics of this market is a submarket of its own, currently dominated by human scalpers as well as a large number of automated computer algorithms programmed by former human scalpers. The competition is just as fierce in this submarket.

Every month, the ECN systems and electronic exchanges change their rates for both fees and rebates, but the dynamic rarely changes drastically. The mainstay names include Arca, NASDAQ (the one that was formerly Island ECN, not the company's other ECN properties), BATS, and Direct Edge's EDGX. Few of the inverse pricing model ECNs still pay a rebate to remove liquidity, but some offer low-priced alternatives and can make for excellent choices to remove liquidity from if there are displayed shares still posted on them when you need to remove liquidity for any reason—either to enter aggressively or to exit at breakeven after an aggressive exit.

Most of the Swifties developed different combinations of strategies after learning these basic dynamics. Some specialized in rebate trading, others specialized in micro-term trending trades (meaning up to two pennies of profit, which was common on a stock like GE back when I began training). It's also important that the style you choose to scalp should match the personality of the stock you trade. This personality tends to be a chicken-and-egg effect resulting from high average volume on a stock causing more scalpers (both human and automated) to scalp it; and then causing others to do so as well, which justifies even more parties to join in. Over the years, many of the high-volume low-priced stocks were popular for slow-moving but easy-to-read movements to scalp one penny at a time, such as Citigroup between 2008 and 2011. Meanwhile, high volume but higher priced stocks tend to be popular for scalpers who aim for more than one penny at a time, such as Bank of America as of this writing, and many mainstays such as General Electric.

High volume scalping in itself is not directly a trading strategy. However, the many strategies that one can develop by simply learning the basic dynamics of this micro-term time frame fall within a small group, in most cases. Some scalpers enter passively with orders that add liquidity and then manage the resulting positions by exiting aggressively. Other scalpers enter aggressively when the odds look good and then try to either exit passively by adding liquidity or exit aggressively by removing on a well-priced ECN.

Many advanced scalpers simply layer orders around a range of price levels to continually collect rebates for adding liquidity as the net position continually changes when their orders are filled. Depending on the average daily range on a stock, and your current maximum trade size, this strategy and its variations are some of the most time tested methods of scalping with ECN rebates used directly in maintaining a positive edge. The exposure to Black Swan risk, however, is also the highest, so many traders who do this as a staple strategy would often benefit by adding another style of trading for strategy diversification as well as a hedge against sudden above-average range days—perhaps caused by major news, major institutional buying or selling, index rebalancing, and market crashes—involving the stock.

During my early experiences with scalping, a senior trader, Ron, stated during lunch hour that any longer term strategies are just dreaming and what he had been looking at was reality. While I wouldn't quite go as far as he did in claiming that trading on the micro-term time frame is the only true reality of the market, it remains true that traders who were brought up in this industry with exposure only to technical analysis charts should at least learn about this aspect of the market. At the very least, next time when you see a slow-moving stock move up or down by one penny on your chart, you would at least understand how much happened in the process (and how much money was made and lost by someone participating in the same market on a different time horizon than you) when you saw a price movement that looked insignificant. Just as children are forced to learn basic math despite the existence of calculators, computers, and cell phone calculator apps, traders of all time frames should at the very least understand the dynamics of the micro-term time frame.

When financial news programs and magazine articles about the securities industry speak of Direct Edge changing their pricing model, or the relaunch of Philadelphia Exchange as NASDAQ OMX PSX with a price-size priority model, no professional trader—whether institutional or individual—should stare blankly as if the article was spoken or written in a foreign language. It would also be constructive if the authors of articles about such news items learned what it meant rather than robotically reciting it with wording that blatantly shows how little the author actually understood the new development.

At the very least, if you're opening a new long-term position for your investment account or a position trading account at a direct access brokerage (and there are quite a few who now pass-through ECN rebates if even just in part), there would be countless opportunities within every typical trading day to passively scale into such a position by posting passive orders that add liquidity or in some other way take advantage of a form of rebate or discount depending on current pricing. It would be worth understanding why you should time your entry to collect a rebate rather than unnecessarily donating money to an ECN simply because you don't understand the meaning of liquidity removal. Surely, there are more worthy causes to donate your money to if you choose to donate your money so generously.