Chasing Unicorns

I spend the later parts of Augusts in the middle of the Black Rock desert, miles from civilization at the greatest party in the universe: Burning Man. I went from not thinking it was going to be my thing to 50% owner in a 36’ RV, purposefully built for the playa, heretofore known as Built 2 Burn (B2B). B2B is currently in the rental pool for Cochella.

B2B is a 1987 Fleetwood Bounder with a non-EFI 454 big block that after a fresh tune-up involving a carb rebuild, cap, rotor, plugs, wires, fuel pump, filters, etc. nets 3 miles per gallon. 4 MPG coming down from Tahoe with a tailwind. We’re currently looking for our next RV to convert into our ultimate playa machine: Burn Force 1. It will be a bad ass diesel pusher for driving up and down the west coast since I’m digging the RV life a lot more than I thought I would.

Touching briefly on some of the comments I saw yesterday, some of the discretionary traders are hung up at picking direction. If you’re an Aim And Shoot chart trader, direction should have simple mechanical rules. Leave the discretionary part to picking entries and exits.

From my past few posts, I’ve detailed my foray into trading. I was lured by the siren into all of the major gurus you’ll find on CNBC, Google and YouTube. I’ve probably taken all of their courses or subscribed to all of their services. Needless to say, most of these guys are peddling garbage.

In late 2015, Quantopian was breaking large on the scene and I started focusing my attention there. Quantopian’s algo development environment and tutorials run on python. There was one problem: I didn’t know python. By then I was fluent in EasyLanguage and knew how to code up just about anything I could imagine in MultiCharts.

Enter my savior, Quant Jesus Dr. Ernie MF’n Chan. Ernie was a quant Rambo, a trail blazer in the quant retail world. Ernie is a PhD, worked on IBM Watson and really fucking smart. But you can tell he’s tried all of the bullshit many of us have, which makes him much more relatable to the aspiring quant. He is arguably responsible for making the word ‘quant’ and quant strategies in general something most people are familiar with.

Most notably, Ernie is a REAL trader! He runs a REAL hedge fund! No theoretical non-sense.

His influence runs deep, which you probably don’t see much in the retail world. Recently, I’ve been helping another trader convert a pair trading strategy from Matlab to run in python. A pair trading strategy in Matlab is Ernie’s calling card, nobody was doing that before Ernie came along. I’ve been helping him retool it to run in pure python.

If this post is worthy of Ernie’s eyeballs, if you do anything Ernie, I think you would 10x your readership if you would teach in python instead of Matlab.

I liked Ernie’s approach because he was testing the shit out of everything he did. There was no voodoo, everything was well tested with detailed explanations of the mechanics of why things worked and what the edge was.

I signed up for a couple of classes at Quantcon taught in python, but I still didn’t know python. It’s a real programming language, it’s object oriented and I had no clue what that meant. So I started on Learn Python The Hard Way and made it through it in about 15 days. I put an ad on Upwork for a python tutor that knew trading and Quantopian.

Here I found Ed, a real python engineer. Ed now works for my firm and is primarily responsible for turning me from a clueless python idiot to object oriented python pro implementing TensorFlow in an hour. With a couple of Skype sessions that involved helping me convert MultiCharts strategies to run on Quantopian, I was digging the python thing and learning it quickly. Quickly enough that within two months, I was comfortable enough to take an advanced workshops taught in python.

I bought a ticket to Quantcon 2016 in NYC and hit the first featured speech: Ernie speaking on volatility. Over the course of the rest of Quantcon, I learned about machine learning, quant bond trading and met a couple of 7 figure producing traders when I was having drinks with their broker.

You mean they don’t use the Shark Fin Indicator to make 7 figures like Wendy Kirkland? My world is ruined.

I also took a bunch of quant workshops, devoured everything in the Quantopian tutorials and regularly combed through the discussion forums on Quantopian. Most of their algos are geared towards hedged strategies moving tons of dollars, which means it’s not geared for day traders. But the astute trader can find enough gold there to apply it in day trading contexts.

I remember the moment I was enlightened in the trading game. I was at a workshop that Ernie was running, of the 13 people there, I was the only one (at the time) who wasn’t an institutional trader. During a break, there was a discussion on how they were all excited about earnings seasons and crushing it with essentially the same strategy, except this was nothing I had heard of before. The way most retail traders trade earnings with options:

1. Trying to guess direction, with no clue about impending IV crush
2. Trying to guess direction, with a mild clue about impending IV crush
3. The semi-pro, direction neutral playing IV crush

These people weren’t doing any of these things. Their strategies hinged around people trying these three things. Duh! It’s like betting on whether or not people would gamble on earnings and it’s a sure thing 99.9999999% of the time.

I feel like I had just been hit by lightning and had seen the light: I’d been chasing unicorns. There was no book, no course, no workshop, no linear algebra refresher that was going to give me the secret I was missing to make everything work. The education I was seeking WAS NOT on Google, YouTube, CNBC or a guru’s chat room.

More importantly, I was learning about actual concepts that WORKED in the professional circles, not retail circles. This changed everything.

No one was going to hand me the keys to the kingdom, I was going to have to work for it and blaze my own trail. No one was going to sell me an automated strategy that netted .3R expectancy and made 100 trades a month. I was going to have to figure it out on my own. With my new grasp of reality, everything changed.

• There were concepts that flat out worked and those that didn’t.
• Strategies that work well long term take 500 hours minimum of R&D time, 1000s of hours long term with Kaizen principals.

The good news is I had taken so many workshops, seminars, read so many books that I had a massive strategy junk yard, similar to that floating chunk of plastic in the ocean. I return to this pile and use these like parts cars – pull the equity curve thing out of this one, pull the position sizing algo out of this one and put it together with everything else I know that works into something unique.

How did I end up as a prop trader? Almost three years ago, I was working from a crowded coffee shop one morning and two traders sat down next to me. One of them was a recruiter for a prop firm, the other was a potential new trader coming from a hedge fund. I was deeply intrigued. I chatted them up and the rest is history.  

So in summary, if Ernie Chan and Euan Sinclair are at a trading expo, I’d buy a ticket and go network. If the CNBC Fast Money chodes are there, I would run in the other direction.

That’s it for today. In my next appearance, I’ll rap about the volatility ETFs and talk about where I would focus my trader education in retrospect.