Tape Reading

I stare at the quotes every day for like eight hours, so I am declaring myself an authority on the subject of tape reading. It’s a misunderstood subject and there’s a lot of interesting stuff I want to go over, including several trade breakdowns that included a clearly defined tape reading edge. There’s also a lot of bullshit to dispel. Some people “think” they’re reading the tape and they’re actually just completely lost in noise. There’s also some cool history about it. Let’s cover the basics first.

What is tape reading?

There are two general meanings when it comes to tape reading. The Tape is often just used as a short-hand metaphor for the general price action of whatever tradeable asset class one is looking at. You go on CNBC and the Fast Money guys are saying “it was a weak tape into the close”, they are just saying the price closed weak relative to the day’s range. Or maybe it closed weak despite “good” manufacturing numbers or CPI.

Then there’s the second more literal meaning of reading the tape, which is to read the order flow in the stock–the buying and selling shit. This originated from Jesse Livermore (he’s the Babe Ruth of trading, go read Reminiscences of a Stock Operator). He literally read the ticker tape in the 1920s, infered the direction in which institutional actors were buying or selling, and made his trades based off of that.

“A battle goes on in the stock market and the tape is your telescope.  You can depend upon it seven out of ten cases.”

In our modern age, reading the tape is now reading an electronic screen, using the Level II and the Time and Sales. You see all the open orders in the electronic market (the level II) and all the transactions that occurred in the market (the time and sales).

SOES Bandits: The Golden Age of Electronic Day Trading

If we’re going to talk about modern tape reading, we have to talk about Datek Securities and the SOES Bandits. I always found this era of trading fascinating because you could do super basic things without any understanding of technicals or fundamentals and still make shit-tons of money. Traders weren’t making money by taking position risk in the market, it’s more apt to say they made money exploiting the market. if you want to read more about these guys, check out Dark Pools by Scott Patterson.

In 1984, the NASDAQ created the Small Order Execution System (SOES).  SOES was an automated trading system that provided immediate execution of small orders entered into the Nasdaq system at the best available price. SOES was the only Nasdaq system that allowed for automatic execution–no phone necessary–and it was also the fastest route. It was rarely used at first but NASDAQ made it mandatory after Black Monday, in response to ordinary mom and pop investors being shut out from trading during the crash. All market makers were required to buy and sell at their quoted price to 1000 shares. That means that any order 1000 shares or less would be automatically executed without market maker discretion. Ping. It’s done. You can’t stall to let a whale at Salomon Brothers or Morgan Stanley take the shares. The playing field was now 100% fair for both retail investors and institutions and there would be absolutely zero unintended consequences to spawn from this. The end.

JK that didn’t happen. This led to the arrival of the SOES Bandits.

Enter Harvey Houtkin and Sheldon Maschler, two stock brokers from no-name boutique shops who had their clocks cleaned on Black Monday. Houtkin observed a lag time between quote shifts that could be exploited via SOES. This was simple order flow arbitrage.

When the market for a stock started to shift—say, offers to sell Microsoft jumped from $50 to $50 2/4;, while the buy offers rose from $49¾ to $50¼—some market makers might be caught napping, still offering to sell at $50. That meant that for a brief moment, free money was on the table. A market maker was offering to sell Microsoft for $50 even as another was offering to buy for $50¼. With a few keystrokes, Houtkin could rapidly snap up a thousand shares of Microsoft for $50 a piece from the slowpoke market maker. Seconds later, he could turn around and unload it for $50¼, an instant profit of twenty-five cents a share. For a thousand-share order, that added up to a quick hit of $250. Rinse, repeat. Dozens of such trades a day added up to real money. Since the trades were typically a thousand shares, the numbers weren’t insignificant: at $50 a piece, a thousand shares of Microsoft cost $50,000. Houtkin didn’t realize it, but he was mimicking a trading strategy, perfected on the floors of Chicago’s futures trading pits, called scalping. It was in many ways the same spread-capturing strategy that Nasdaq market makers used. The difference was that market makers made money from buying and selling to regular investors. Scalpers like Houtkin made money off the market makers.  

Nasdaq market makers hated it. The $250 profit for Houtkin amounted to a $250 loss on their end, since they could have sold the stock for the higher price or bought for a lower price. SOES forced market makers to become more diligent about the stocks they juggled. It wasn’t easy. They often handled twenty or thirty stocks at a time. Without a sophisticated computer system, it was nearly impossible for them to keep track of all their stocks at once. Houtkin and Maschler could focus on just a few stocks and whack them time and again. Outraged, pockets picked, the market makers starting calling Houtkin and Maschler the “SOES bandits.”

Houtkin relayed the strategy to Sheldon Maschler (what a character he was) and then taught other traders at Datek, where they would trade even more aggressively than Houtkin did. Maschler didn’t recruit Ivy Leaguers, Chicago School economists, or haughty MBAs. He recruited city college kids from outer boroughs and Long Island. You weren’t supposed to think too hard about how you made money. Maschler’s abrasive style rubbed Houtkin the wrong way and they eventually parted ways. Maschler would pair up with Jeff Citron and together they saw that the future of trading would depend upon having superior technology.

Enter the boy wunderkind, Joshua Levine. To make a Formula One anaology, if traders were the drivers looking to race through fast market conditions, Josh would be their chief engineer–the guy who oversees the design of the car. In 1989 Levine created Watcher, a trading interface that would quote and execute faster than the old-ass, piece of shit they used to use, Nasdaq Workstation II. It ran faster and would also display specific trading opportunities such as lagging quotes from slow market makers. He created something called the Monster Key that allowed Datek traders to capitalize on NASDAQ’s price-time priority rule by adding a 20% price increase to all limit orders. If you knew how to make money, Levine would give you the best tools. Levine also invented one of the first ECNs called “Island”, later acquired by Instinet. His fingerprints are all over the architecture of the modern market, even to this day.

Using Watcher and trading on SOES, traders just had to find volatile stocks and pick off slow quotes and slow market makers. MSCO would be on the offer for 1000 shares, you take it and you were instantly in the money. if you knew absolutely nothing, you could protect your gains and flip the stock quickly. If you had a little more experience on how far certain names, like AMZN or YHOO, could go in the tech bubble, you could hold on a little longer. Compare that to today’s market with HFTs where you can rarely find scalping opportunities. You have to sit and wait, position trade, and set wider stops. Maybe you endure choppy whipsaw price action for the entire day. It fucking sucks.

An Average SOES trade breakdown

Ok let’s breakdown an average SOES trade in 2000. No, I wasn’t there but I’m going to pretend I was. I’m pretty sure I would’ve made a shit-ton of money. This is entirely hypothetical, don’t expect historical accuracy. I’m trading a dotcom stock. This World Wide Web thing is pretty cool and these dotcom stocks will take over the world. 

It’s 9:30. Yahoo just announced they bought a company. They’re doing something revolutionary, an absolute banger of an acquisition that will change the economical paradigm forever. It’s Radio on the Internet. Holy shit. There’s all this buzz on the prop desk, you can hear the murmurs in deep Long-Island accents. Let’s just bring up YHOO on the level II and buy it. 

Oh that’s great. Some mook at Morgan Stanley is still offering below the bid (stale quote). I’m going to rip him for 1000 shares. I get my fill. I’m already a point and a half in the money for. I could flip it back to him for $1500, right then and there! Free money!

But I ain’t gonna do that. $1500 ain’t shit. Shelly and Jeff might be right behind me, ready slap me upside the head for being such a piker. I know Radio on Internet is a huge deal and they’re going to walk up the stock to 120? 130? Doesn’t matter the price–I just know stocks only go up. It’s year 2000 and this World Wide Web thing is pretty cool and these dotcom stocks will take over the world. 

Two minutes later.

Seems like everyone woke up finally. They are clustering the bid and pounding the offer aggressively. No can get stock fast enough.

I am now 5 and a half points in the money for $5500. I’ve taken zero pain. There is zero noise because we are still a year away from decimalization and seven years away from the first appearance of HFTs. These market makers are just walking the stock up. Some of the traders here who didn’t get the opening quote are still buying anyway. I’m going to just sell when they start to sell it, that’s it. If they just keep walking it up to the moon, I’ll ride it with them.

Ten minutes later.

Every market maker still on the bid to close the spread. There’s a little more supply but we’re on a rampage. Still NO REASON TO SELL.

Another ten minutes later. It’s not even 10am yet.

The initial opening move has slowed down. We’re not printing the offer as much. The bid is thinning out and ready to drop. The offer is starting to cluster. The open has probably found a climax and I want to lock in my $18,000 profit. Good enough. I click sell and fill my 1000 shares and then the stock immediately takes a down tick. There’s a pullback to the low 120s. I go out for a smoke break. I chat with some of the guys about how great this job is.

30 minutes later I come back. YHOO is at $150. There’s some chatter that Jeff Citron is up over six figures and hasn’t even sold yet. Fuck me… I need to quit smoking.

It eventually ends (it always does)

The Golden Age would eventually end. It would never be that easy again. The dotcom bubble busted. Volatility left the market. SOES would eventually get phased out. Decimalizliation, hybrid market and HFT would make the method of pure tape reading to make money a relic of the past. The prop industry shrunk big time and consolidated into just a few firms. Traders today don’t really read the tape to make money, they have to take position risks and be right about their ideas. They have to sweat out the swings rather than be instantly in the money upon entry.

Tape edges in my life:

Even though I started trading in 2010, a full decade past the dotcom bubble, I was fortunate enough to have some dicovered some of my own order flow exploits. It’s nice to hack the market to make money versus actually taking position risk. It’s rip and run trading, so straight forward and so simple. I miss it. Sitting through small cap short positions and seeing the most fucked up squeezes go against you? Ugh, I can feel the bile rising in my throat just thinking about it.

1. Tradenet “one cent risk” trade — CFD “cheat code trading” on the uptick rule
2. The KODK trade — timing the LUDP 5-second timeframe on extreme margin-call driven volatility halts
3. FNMA/AAMRQ/Weed stocks/GBTC — exact bottom and tops to the exact penny (or thousandth of a penny) on OTC
4. The ARCAEDGE pre-market trade — taking advatange of a fragmented market

All of these methods made a lot of money for me but they eventually lost their efficacy, which is why I’m writing about them now. I broke down the fourth one quite awhile ago and will breakdown each one in a new post, provided I still feel like blogging.