Things You Learn After 1 Year of Day Trading for a Living
I’ve made 20% ROI in 6 months and lost it all in a single month. Then I lost 30% in 10 trades the following month. I’ve earned and lost thousands of dollars, almost blew up my entire account twice, got dozens of margin calls, tried multiple Machine Learning techniques, traded multiple markets, time-frames and instruments. I’ve made any possible mistake, but somehow I survived and learned a lot.
After 4 years in the Software Engineering industry, I realized my path was too predictable. I would always deal with Data Science related projects. Working in a small company, enterprise and a startup shaped my industry perspective but nothing was quite satisfying. My good old passion for Algorithmic Trading would never leave me alone. I wanted something else, so I decided to quit my Data Science career and pursue day trading for a living. Here is summary of mistakes I’ve made and how you can avoid them.
We have seen Machine Learning applications everywhere. The media loves it, people don’t understand it and investors call it “buzz words” because they will never get it. DeepMind’s AlphaGo progress is mind-blowing, and it’s only the beginning. I was always telling my colleagues that someday we will sit and think about Java classes as they appear in Eclipse in front of our eyes, and during our sleep time we will be injected with some books “read” like in The Matrix.
The problem with Machine Learning is that it’s very tough to apply in trading. It’s more of a filtering method rather than a decision making tool. Most of the paper trading tests will be awesome and will fail in real trading because they over-fit. You will fight it with cross validation and cherry pick the best models that performed best on out of sample, thinking you are safe, in a way adding bias and leaking data. This is not the way to do that.
Avoid over-fitting by carefully averaging and evaluating on different assets, time frames or periods. Use non-conventional train/test splits and add random noise to evaluate your generalization power. Always be extremely careful, because you don’t know what you don’t know. You don’t know what’s going to happen, but Monte-Carlo simulation is your best friend, as you can at least simulate a lot of scenarios.
Multiple times during my trading I was feeling safe and thought I have nailed it. I felt like there is nothing that can surprise me, and time after time I was slapped in my face by mister market. Single trade going wrong and wiping out your previous 10 profitable trades, volatility spikes and your stops are breached like paper by knifes and free liquidity turns into a killing strangle on your portfolio, following a nasty margin call and your broker fixing all of your positions before it’s a total loss. Believe me, unless you saw that, you think it’s just war stories and imagination. Folks, this is reality, there is no free money out there. Everything involves risk, it’s a matter of how good you are at understanding your odds and your probabilities of loss.
As an options trader, my edge relies on selling overpriced options and buying them back when prices drop. Several pricing models exist nowadays such as Black-Scholes-Merton, Binomial Trees and Monte Carlo. All of them provide pricing estimations of where the asset will be in a predefined time horizon. Usually IV (Implied Volatility) overstates the fear in the marketplace. But sometimes the fear is real.
As we have seen in February 2018, market fear is sometimes real. Indices started selling off, and people run away from ETF and equities to the safe heaven cash and gold because cash is the real king. Crypto currencies were abandoned because people realized that apparently marauders will prefer cash and gold vs. digital money.
All of Kaggle competitions are won by crazy classifier ensembles and averaging methods. I tried to be a smart guy for a long time by applying cutting edge techniques, algorithms and tools. They work sometime, and sometimes they don’t. Usually, it will take you weeks or months to understand what went wrong. This evaluation costs you money, or you paper trade it aside the market, and as mentioned before, this is a non-deterministic process that just adds noise and leaks data.
The truth is, simple statistics, Monte Carlo simulation and a little bit of Python is all you need. Simple assumptions well simulated and validated can spot over-priced and under-priced bids and asks, that’s it. The fancy models are good for your ego and general understanding. In markets, a CS 101 probability and statistics is good enough for a profitable strategy.
If somebody with no trading experience asks you how you make money, you must be able to explain it in couple of sentences, otherwise, you are not making money. Prior to that I used to explain people how fancy my Machine Learning flows are, without being able to explain the alpha. Now it’s simple: I sell overpriced options premium, that’s it. The asset price bounding methodology is pretty complex, but the alpha source is clear.
Trading seems like a difficult task for most people, which requires training and financial education as a prerequisite. The truth is people act as traders each and every day without even noticing. You go to the supermarket to buy stuff. Here is your market (exchange). Prices are your bid-ask-spreads (level 1). You can only buy (ask), but the supplier can also sell (bid). The supply at the back of the supermarket is level 2. The cashier is your order-book. Now extrapolate it to everything you buy or sell, it’s all trading, everybody is a day trader.
Seriously, the more complexity I was adding to my algos, the larger were my losses. Clouded by fancy Spark jobs, Lambda expressions and beautiful Jupyter notebooks, I actually was making less money. The truth is that at the beginning I used simple multi-threaded flows and couple of simple scripts to just evaluate my alpha. The moment I began concentrating on performance and ease, I lost track of the alpha itself. The most important thing is to keep track of a simple and working flow, then you can add the jewelry, on top of a strong skeleton. Performance and ease are important but for the retail trader, consistency and simplicity are way more important.
Being profitable for 6 months is nice, but you can always lose more than the couple of previous months. Sometimes the market is brutal and fast like an alligator. Avoid those situations by playing small. One of the biggest mistakes I made was over betting. Although Kelly criterion is important consideration, under betting is always better than over betting. Risk assessments and position sizing are key to your success. Having a strategy with high probability of winning is as important as correct position sizing and margin requirements analysis.
People will tell you what should have been done constantly. Every mistake I made was followed by someone telling me it could have been avoided. Your family, friends and colleagues will doubt you, your alpha, your skills and your ideas. It seems like Robert Kiyosaki was extremely right: people are too afraid to try stuff, so it’s always easier to evaluate somebody else, instead of doing it. Most of the time you will be discouraged by your surrounding environment: from trolls on social nets to your potential investors, everybody think it’s easy, and they can do that, but then again they are just watching and mastering the art of hindsight. Follow your instincts and gut feeling, you are in the arena, fighting the battle, don’t get mislead by the spectators.
That’s going to happen to you too. Most of the pro trades specify the psychological robustness needed for the game. In times of volatility moves you will feel your heartbeat as market opens, margin requirements will suddenly jump from 40% free to 2% prior to market open. You will be hopeless and miserable, because you made very large trades with unbounded risk, and didn’t think about the rainy day. Some days will be rainy, always be prepared.
Moreover, I lost my soul. I got repetitive comments from my friends about being “cold”, “aggressive” and “rude”. In some moment I almost forgot how to play the guitar. Every social event was suddenly annoying and time consuming, or a waste of precious coding time to me. Long working hours and weekends full of development and hundreds of commits, eating disorders and the most obvious loss of weight. This is extreme, don’t do that. I read somewhere it was actually a sign of doing a great job in my endeavor. It’s bad for your trading. You must avoid it, don’t exaggerate.
After messing up time after time, I’ve drastically reduced my expectations. Suddenly I understood the well-known saying regarding how much money were you able to actually take and keep from the markets. Living out of savings, I’ve realized I’ve got gas for couple more months, and it’s time to pay myself a monthly salary, otherwise I’ve done nothing.
The most important thing is that suddenly I was fearless, nothing could frighten me anymore. I wasn’t naive anymore and understood the risks involved completely. In a way I realized how fragile and dangerous this business is. It’s like walking in a minefield.
I started trading small, really small. Instead of jumping into trades like a panther, I was investigating the company first, plus usually multiple trade ideas will appear for the same symbol, so there is no FoMO (Fear of Missing Out). Moreover I reduced my watch-lists significantly, focusing on liquidity and volume. This a-ha moment was the most significant. Getting in and out of a trade is mandatory. All of my losing trades were with low liquidity assets and bad fundamentals, things that takes you seconds to evaluate nowadays.
Before my major losses, I used to make 4% ROC a month. Commissions seemed irrelevant and minor. The moment I blew up more than 30% of my account and reduced my positions, I suddenly realized the importance of low commissions. Trading with IB (Interactive Brokers) I do pay lowest commissions in the industry, considering the robust platform and flexible API. On the other hand, by reducing my position sizing, I found myself paying 5–10% commissions and IB suggested to “provide liquidity” as they “do not negotiate rates”, even if you make 200 trades a month. Moreover I need historical EOD (End of Day) prices so Quandl is vital here. Brokers don’t give a duck about you, being a retail trader sucks.
The only way to avoid commission ripping is trading size. People tend to talk about diversification and all of that stuff. How positions should be small and so on. Well the reality shows that trading too small kills you. This is a personal parameter and a function of your account size, risk aversion etc.
One of the toughest things to accomplish during day trading is patience. Sometimes the best trade is not to trade, similar to Zugzwang in chess. Being a day trader means being a market junkie, which implies addiction and adrenaline rush during the opening bell. Sometimes cash is king, simply not trading or waiting for the appropriate trade is the best you can do, especially when markets go crazy in times of sellouts or crisis. On the other hand trading is like air, if you don’t trade you do not exist. Mastering this urge is key to your success. So many times I have been adding to losing positions or trying to save terminal positions, instead of waiting and keeping the cash.
Patience is also relevant to entry and exits. You will see prices you want, but then they will change the LMT order, and you didn’t get it. Then you will adjust and chase the price which will move again. Don’t rush and be strong, it takes time. Eventually you will hold on to your opinions and wait for the other side to take it. From my experience if the underlying is liquid, all day trades with middle prices will be filled.
If you have a losing trade, there is no escape and expiration is weekly, don’t rush to fix it. The market can bounce, and you will be naked. Your losses could get smaller. I had a bear spread after the market selloff in Feb 2018, fixed it with 0.7 loss of the spread width. The next day it became 0.2, so waiting sometimes helps. Never keep your losing positions naked, markets can always bounce, even if it’s one day till expiration.
Trading is definitely more art than science. You may start playing a new instrument right away and probably anyone could do some sounds after a weeks or so. After a month you will be able to play some notes and hopefully a song. The question is how long will it take you to play like Steve Vai? He mentioned practicing 8 hours a day, and sure he is gifted, but then again, hard work is key.
Similarly, trading requires a lot of practice. Getting to a level of trading effortlessly is what divides professionals and hobby traders. Unfortunately, people don’t get this concept because day traders always look like lucky cowboys with money to spend. The path is what people seem to omit as it’s the most difficult part of this game.
I’ve read dozens of books, hundreds of articles and watched hundreds of hours of video related content. The learning never stops. Every day you don’t learn something new is a lost day. Your trading plan must be improved and engineered constantly, otherwise you won’t be relevant and your edge will disappear. The best content is available online and mainly for free. If you are into options trading, you must check out tastytrade, OptionAlpha and OIC. Those guys will teach you everything you need to know. It’s really a matter of how motivated you are.
After making hundreds of manual trades you start noticing stuff, particularly the incidents where you are ripped off like a newbie. One of the most frustrating concepts in trading options, besides the commissions is market makers. Market makers are essentially the players that run the show. It’s widely known that companies like Citadel, Final and a lot of other HFT (High Frequency Trading) players completely control the price inefficiencies in the markets. Big institutions trying to buy or sell are main targets for ripping off, because fast players can easily get before you and sell you the same stuff in higher price or buy it from you in a lower price, and then sell or buy it to/from someone else. Those minor differences compound like a snow ball.
As mentioned before, commissions are part of the problem, but without them there will be no arenas to trade in. The real issue is market makers bluffing the order books. All of the infrastructures are automated, and the fast players are everywhere to catch your trades, happily providing you high prices when buying and low prices when selling. The only way to beat it is to use limit orders and try to anticipate the middle price. This a-ha moment seems like a minor issue, but multiplying 200 trades by 2.5$ ripping off fees is similar to the commissions you pay anyway, so it’s like paying commissions twice! Never use market orders or bid-ask raw prices, always target the mid-price or better.
Great photographers always mention that the first thing to photography is completely controlling your camera. You must understand and know how to utilize any setup and combination depending on lighting conditions. Once you know it, then the real art begins.
It took me 6 months to fully utilize my trading software and use the API effortlessly. I have a full clone of my trading setup on my PC, laptop and an AWS instance. In case of failure I can easily resume my trading immediately with all the software I need. My network connectivity is pretty stable. On bad weather or rare incidents I have multiple network adapters so that my smartphone becomes a hot spot. I added multiple automation layers to make my trading robust and consistent as possible. Analysis paralysis is bad, particularly in trading. Try to eliminate manual interrogations as much as you can.
I made dozens of fat finger errors, and probably lost couple of thousands due to wrong prices and combinations of multiple trades. Rushing and lack of knowledge will lead to dumb mistakes and loss of capital. Don’t run, you are a retail trader, not an algorithmic trading computer. Learn your trading software thoroughly.
Over-trading is bad. Trading is super exciting and you become a junkie. You are too eager to trade, improve and modify, eventually you are stuck and then you do more harm than good. At the beginning I wasn’t able to leave my screens. I had futures and tastytrade broadcast on one screen, and my positions on the other screen. Eventually I would sit for 6 hours and modify, trade, improve or generally ‘touch’ my trading. I learned the hard way that trading options is done at the opening bell and closing bells only. Everything else is bad.
Removing balance, PNL market value and all money related indicators of my portfolio is good. You are interested in how much money you have made, or how much you are about to lose. That’s going to kill you. Dropping money from the routine is good for your performance. You must think in probabilities and risk to reward rather than in dollars. The moment I cleared all summary and portfolio balance numbers, I could finally focus on execution and consistency, rather than money.
The only way to survive in this game is to trade like a robot. Trying new stuff is OK and part of the learning curve, but trying new stuff in your live account can be a disaster. I was switching my probability of profit thresholds and my risk to reward ratios too fast. Eventually I was able to converge and find my optimal ratios. Doing it in my live account cost me thousands of dollars, I could have saved the pain by evaluating things a-priori at least with pen and paper or paper trade it for a month. Being consistent and persistent is mandatory.
PDT (Pattern Day Trader) rule requires a minimum of 25K$ to day trade: make more than 3 day trades a week. I hate it, everybody hates it and think it’s stupid. It’s annoying but the logic is sound once you are forced the PDT restriction after trading like a cowboy.
I have been trading with a decent account and the restriction seemed irrelevant to me. I never had to actually prioritize my trades, as I could make them all. The moment I lost half of my account, I suddenly realized how precious each and every trade was. Instead of shooting all over, I had to laser focus my trades.
I was interested to do some statistical analysis of my trades, particularly the losing ones. Let us divide the analysis in terms of raw trades vs. symbols traded. This minor difference ended up being very important.
Since June 2017 till March 2018 I’ve made 1068 trades (621 profitable and 447 losing), that’s 0.58 probability for a win and 0.42 for a loss. I’ve earned 87459$ and lost 122069$. The average win is 140$ and average loss is 273$. That’s a 0.51 return to risk ratio. The effective edge is defined as following.
Of course there is no edge due to the low probability of profit and high risk to reward ratio. Let us now group the trades by symbols. I’ve traded 130 symbols with positive PNL and 51 symbols with a negative PNL, a total of 181 symbols. 0.72 of the symbols were profitable and 0.28 were losing. We have an average 134$ profit and average 1020$ loss for each symbol, that’s a 0.13 return to risk ratio. The effective edge is defined as following.
Again there is no edge and this is even worse. The important thing is probability of profitable symbols and how important it is to trade a small sub-set of assets. Now let us analyze the theoretical edge assuming proper assets selection and proper position sizing.
16% it is. I just proved to myself that trading small and often is key to success. Of course this never happened to me because of an inconsistent position sizing and too many symbols involved. The moment you tilt your trades, you are doomed. Taking more than 1 to 2 return to risk is a losing game. This one was probably the largest a-ha moment to me.
You hear a lot about how trading journals are important, but honestly, nobody keeps one. You think it’s useless because you see your trades anyway in the trading platform and in general you think it’s a waste of time. You are only interested in your winnings and how much money you make. Eventually you will have to grow up as a trader, and you will realize how important the trading journal is.
Keeping an up to date trading journal will improve everything. From commissions and odds to assets you trade. You will learn more than you think, and will differently improve your discipline. I started running a Google Sheet as a trading journal. One week after running the journal I realized my risk was too high and my trades were too small. Effectively I was risking way more than 1 to 4, the reality was close to 1 to 5 because my trades were too small.
For instance in my options strategies I was usually selling at least 0.5 credit, or 50$, and then analyzing the profits you realize at best you actually make 0.19 or 19$ (options contracts have a standard 100 multiplier). So let’s say you are risking 100$ to make 25$ which is 1 to 4 and the probabilities are in your favor, but in reality you make 19$ and risk the same 100$. The drop from 25$ to 19$ is a combination of widely suggested fix at 50% plus the courtesy of the broker taking between 3$ and 10$ per trade round trip (in and out). Again those minor differences compound like a snow ball, and reduce your edge.
The only solution to this problem is raising your minimum entry price. From that moment I’m taking no less than 100$ premium. Instead of 50% profit taker, I use 75%. Effectively, we have 75$ to 200$ risk which is 1 to 2.67. This ratio is bad but realistic. With 75% winning trades we are OK, and it’s easier than 80% to get. Although it’s common to use long distance options strategies, the weeklies work for me, so you will have to adjust your profit takers and stop loss levels accordingly.
SPX@CBOE or S&P 500 Index is the market. Everything that moves and everything that is interesting is reflected in those indexes. Every crash, peak, hype and fear is there. VIX@CBOE or the Volatility Index of SPX is the pure reflection of fear and greed in the market. When markets move, the volatility moves and vice-versa. Correlations work for the long term, but when volatility spikes, everything is correlated.
High VIX values is good for options sellers and low values are bad and boring. Note that long periods of low VIX end up in massive explosions.
Keep in sight the most moving assets for the day. Assets going down are more interesting as premium is going up. As premium sellers that’s when the action is lighting fast and the bids are juicy. Market fear is good for options trades as premium goes up. Be patient to catch those moments and act immediately. That’s when your edge is wide and clear.
The inability to get a fill for your trades will drive you crazy. Multiple times I was chasing prices until I got it, but did more harm than good. Market makers will always show you a better fill the moment you are in, and will seldom provide you the mid-price or a better fill than was requested. Don’t fight it. Prior to getting in, just find bids that satisfy your risk to reward ratios. Two things will almost always happen. 1. You will get your fill. 2. You will see a better price immediately. Note that this game is unbeatable, but at least you are within your risk to reward. Make a lot of trades and you will be fine. Difficulty to realize that will lead to one of the two: 1. Almost no fills 2. Negative expectancy in terms of risk to reward due to commissions and your target exit price (which is seldom 0).
Unfortunately it’s thought to apply and a clear filter of successful trades vs. losing trades. Remember to check yourself before every trade. Multiple times I’ve done trading mistakes and repeated them again and again.
Remember that trading takes years to master and it’s a tough walk to walk. The learning never stops. Markets are dynamic and alive. Alpha tends to disappear as cars run out of gas. It’s important to reevaluate and refill your strategy base.
Hope this summary will save you time and money. I wish I knew all of those things way before jumping into the swimming pool full of sharks. Be careful as we are small retail traders and the sharks love us fat stupid snacks.
Check out my github: https://github.com/algonell
Feel free to contact me: kreimer.andrew(at)gmail.com
Things You Learn After 1 Year of Day Trading for a Living
Research & References of Things You Learn After 1 Year of Day Trading for a Living|A&C Accounting And Tax Services