Skip to content

Prevail Versus Fail: Sophistication and Preparation

Many traders I come across have a very loosely defined notion of what they intend to do every day: identify some support and resistance, see how price is trending, check the news, run some indicators, check book value, PE Ratios, etc. There is a lack of process and definition behind what they do. And this typically continues for x period of time until they either tap out voluntarily, involuntarily, or wake up from the dream.

The “noise” out there is louder than it has ever been. I open pretty much all social media and news sites these days with “mind earplugs” on, assuming that someone is always trying to convince me of something, whether it be an idea, a perception about their lifestyle, etc. But contrary to my own actions, people are much more easily persuaded by noise than anything else.

It is human nature. Hear a loud noise, you turn your head.

And it is why certain strategies, thought processes, etc., driven by others become our own. It is also why we seem so confused all time, especially when it comes to something like active trading. The noise never shuts off. Few people “return to center” and focus on themselves, if that makes sense.

I recently went on a hunt for other methods of pricing crude oil. A lot has changed in just the couple short years since I created my first version of what I still use today, and thought I might some things that could benefit it, overall. One thing in particular triggered my interest: it was essentially a mishmash of several different traditional model types run through a highly specific learning engine.

I was astounded with the level of accuracy achieved by using such basic inputs. When I looked closer at the results, a lot was generally inconsistent with what most people see on the retail side of things – sparse trading and an overall unique perspective in terms of how to manipulate the data to achieve optimal results.

Lately, I have seen many “bad” examples of this in the cryptocurrency world. One of the things that drew me to it was the wide open access to tons of data. You can create anything you can imagine in terms of models and indeed, many people have. And yet, despite having the world of data at their fingertips, many produce very poor results. Typically, I believe its because the people creating them just simply lack the experience when it comes to markets and pricing. This is not their regime.

As usual, much of the time it does not boil down to what you're using, but rather how you use it.

It's a simple reminder of everything I have seen to date with traders of every size imaginable. Everyone seems to be using the same thing, and yet some prosper, and many fail. Is it the inputs? Interpretation? Many times, yes. But what about the people who have been trading for years, seem to know just about everything, yet can't seem to put it all together properly?

Most of the time, this boils down to process. Before we get too far, here are couple short stories:

Tales of 2 Quants

Renaissance Technologies is one of the most admired, yet secretive hedge funds (in terms of strategy) on the planet. They are known to hire only the brightest of people with proven track records in scientific critical and creative thinking /problem solving, not necessarily finance. To this day, I have only known one employee there (a former boss' boss) and there's basically no way I would ever have an idea in terms of what they're doing behind the scenes. A Quora thread provided some broad-level insights in terms of what they employ on a daily basis:

What I found most interesting here is that every possible angle is observed, and execution is completed with the highest precision. The core strategy on its own (statistical arbitrage) is one which would be considered old, and discussed countless times in all forms of media. It has less to do with what is done vs. the level of sophistication in terms of how its done.

Which brings me to my next short story: In 2008/09 I was working for a large hedge fund, the majority of which was long/short equity. During this period, the worst performing fund  was quantitatively driven. What happened? 2009 was underway, and (from what I was told – last thing I would do as a 26 year old uninvolved would go up to a 50 year old PM and ask him why he was screwing up) option vols went through the roof and were, bottom line, not expected to go so high. The strategy couldn't carry the weight, and it ultimately imploded. Compared to the rest of the firm (which was net bearish at the time) final performance didn't even come close. Many of the book's positions had to be collapsed manually in order to stop the hemorrhaging. Up until this point, the returns were consistent, admirable and overall would have been welcome in a large number of other firms.

These stories demonstrate what appears to be true in just about everything in life.

Put two craftsman in the same workshop, with the same materials, same tools and give them the same amount of time: one will build something incredible and the other will turn into a  sarcastic “nailed it” meme. In this (or any case) the process followed by person A couldn't look anymore different than the one followed by person B.

I have always argued that you can turn a profit or lose money in just about anything, and it is not just trading. Many VC's enter agreements with companies where the bar is set pretty low, hoping that one day they'll hit about 5,000 home runs with one company, reaping returns for the broader portfolio.

Clean Up the Mess

The common path for most traders is one of confusion: learn, learn, learn, and then apply it as if you are throwing a tennis ball soaked in pasta sauce at a white wall.

What you end up with is an uncontrolled mess (and now you have to clean it all up).

Look at the standard assumptions behind what you read, and assume it's all wrong. Again, a large portion of what you read out there is either uneducated noise or purposely malicious. For example, conventional wisdom tells us to wait until a price pattern is completed before we trade them. My issue with these is that the majority are average performers, taken in this sense. A “bullish flag” pattern fails about as much as it follows through, to the point where you might just want to call it a “bearish flag”. But one thing is certain: it, and many others like it, begins the same. Identifying an emerging pattern (or potential group of patterns), and trading into their completion, is an alternative. Whether the flag fails or not, we know the base of it has a higher probability of holding than the completed structure following through in and of itself, based on common assumption.

I see this even more frequently when it comes to all things “order flow”. People have a hard time wrapping their heads around something which is truly counterintuitive. I've found over the years that with trading, things are never what they initially seem (understatement of the century). Limit absorption is a good example of this. You'll see rushes of orders lifting the offer, price goes no where, trader shorts, thinking its going down because of this, then all of a sudden gets plowed because those market orders didn't give up. They're guilty of applying a strategy which may be appropriate in one market to another, which was incompatible with the thesis due to liquidity. Unseen, unobserved, and “naive” as some might call it. Either way, the qualifiers were misinterpreted.

But anybody I know who has been able to survive the gauntlet of trading follows a process, whether they realize it or not. They have clear variables, instances and targets with risk defined before any action occurs, so that even if their underlying assumptions are incorrect, they're still able to stay afloat by way of a repeatable process with appropriate risk.

But easier said than done:

1. Take an inventory of what you have, both tangible and intangible. Forecasting price is about using what we know, and manipulating it in order to make an informed decision. Take your high level ideas (support and resistance, whatever they may be) and break them down into digestible components. Make a list. Some of those items are going serve higher value than others.

2. Organize your list in order of importance. Anytime I'm personally trading I'm really only looking at 4 different, high level items. Each “setup” is based on the first item in each bucket.

Each one of the items in the entry happens very quickly, and oftentimes, everything after point #1 is just padding. Knowing what you're looking for in advance sits on the top tier of the plan. You are defining everything you need to know before you do anything at all. Obviously, you need to have a comprehensive understanding of what you're looking at, in the first place, in order to make an informed decision. I trade a lot of futures so my story is going to look different than many of yours, but the underlying concept remains the same.

3. Identify your setups. Setups are usually determined by market conditions. About 3 years ago now I did a lecture on convergent and divergent trading strategies. Essentially it refers to trend following vs. mean reversion. Bad algorithmic strategies try to trade through everything with zero comprehension of environment or volatility because they're too “dumb” to realize that the same signal winning in one environment will pose significant losses in another. The vast majority of automated strategies I have ever seen hit the internet have no concept of trading environment and all look for the same things: 100,000 variations of an RSI overbought/oversold.

So yes, organize your setups. Be highly specific in terms of what you're seeking and leave zero gray area for when it comes times to making a move.

Want to use support and resistance? Fine. How, specifically, are you defining the levels? What chart resolution are you using to do it? How many historical hits on the level, or do you prefer to use profile?

Want to use volumes? Great. What specific number of executions are you looking for, or are you using a relative percentage or other variable? Where do they need occur? If you're using a series, how many?

And then where, specifically, are you using this? How many ticks for execution? What is your cost per trade? How is your risk defined? Soft or hard exit? What is it and what's the trigger?

4. After all is said and done, for entry, you should have a relatively short list of items you're seeking (for short-term trading). Fundamental valuations can go much deeper. You're welcome to make this list as lengthy as you wish, but bear in mind that humans can only look at one thing at a time. “Multitasking” is basically the speed at which you go from one thing to the next, and how well you retain that information when you do.

Keep it as simple, but as thoughtful, as possible. Everything should add value.

5. Determine what instruments you're going to apply this on. Performance varies due to the nature of each instrument. Every instrument moves in accordance with how saturated it is. More orders = more friction = slower market.

6. Visuals or other tools. What tools need to be loaded or panels in view for you to execute? They should be there/available at all times with zero friction to access.

7. And after all of this is assembled, you need to know how it stands the test of time. There are plenty of platforms these days that allow you to do this on the fly, or with both manual and algorithmic backtesting capabilities. Use them. But don't risk a dime until you know what you're doing. I have heard, thousands of times, the argument about demo vs. live trading and how its not the same because of all the ideas floating around in your head – but that's all after the fact.  There is zero sense in wasting money on something that could turn into futile effort, when you had the tools in the first place. This should be known in advance.

All of this just boils down to being specific.

I worked with a guy who would call every trade imaginable in demo but couldn't execute live. I asked him to walk me through his specific process and he could barely do it. Essentially everything was a pile of ideas swimming around in the back of his brain (which he was right about) but never spit it out. His live execution consisted of very few trades of moderate performance. He explained that he would get in a trade and start questioning everything. Pile of confusion.

So even though your live performance might look a lot different than demo (assuming you're anything like the guy above) it beats the hell out of the alternative. There is no worse feeling than getting into a position and not knowing what's going on, when you do.

Wrapping Up

That's about all for today. Consider this post more of an extended rant in order to get me moving again from a writing perspective. I have been spending a lot of time on Twitter these days, although sometimes the noise and negativity of the things I read on there gets draining. I'll do my best, but we'll see. I stopped writing a lot after moving to the burbs and doing the family thing. Time = sunk (but well spent), but I still make myself available when called upon here.

So until next time, rant over. Thanks,