Wednesday, November 25, 2020
Home Blog

The Super Simple Bitcoin Trading System

0

In late fall of 2017 I created a very simple Bitcoin trading strategy which I have described with all its trading rules here.

I like simplicity and in algorithmic trading, simplicity often trumps sophistication especially when it comes to robustness and how trading models perform in real time. Over 18months have passed (wow, time is flying) since and I want to share the results.

How did a Bitcoin Hodler (buy-and-hold investor) perform from 2018 until end of June?
Well, if you have followed Bitcoin then you know that 2018 was a brutal year for Bitcoin and even worse for alt coins. On the first of January Bitcoin was trading at 13’700$. After trading up to 17’200$ it then plummeted to 3’200$ and only from April first (April fools day?) 2019 Bitcoin was picking up steam again.
Overall a Bitcoin Hodler lost 1425$ per BTC but the drawdown of over 14’000$ would have required nerves of steel.

How did the “Super Simple Bitcoin Trading Strategy” perform (01.01.18 – 30.06.19)?

Remember this is all out-of-sample since the trading model was created in Fall 2017 and has not been changed.
The trading strategy held up nicely returning over 6000$ in profits. More importantly the drawdown was reduced to digestable amount of 3’376$ which is far easier to stomach than beeing down over 14000$.
Bitcoin trading results
Above you can see the equity curve and the corresponding drawdowns. Of course 2018 wasnt a good year for this model either as it traded mainly sidewards beeing under water about 1000$ most of the time. However keep in mind that this trading strategy was designed to try to outperform bitcoin on a buy-and-hold basis and was a long only strategy. It certainly did its job nicely but there are some other advantages:

  • Simplicity: If you read the trading rules you will see it doesnt get much easier than this. Simple is good. The rules can be found here here.
  • It outperformed Bitcoin. Doesnt mean it will keep doing it in the future but its a good sign (Past performance is not indicative of future returns; Please read full disclaimer on the side.)
  • It takes much (not all) emotion out of it. After a nice run up, I dont need to stress when to take profits. I simply get out according to the rules.
  • Because of the nature of the trading system, you will always be on board when bitcoin breaks out, hence its a breakout trading system.
  • With only 20 trades in 18 months, you can easily keep your btc in cold storage 99.9% of the time.
  • The performance would outperform bitcoin buy-n-hold massively with a position sizing algorithm. Think about the reduced drawdowns. If you had 100’000$ in BTC by January 2018 then come April 2019 you would only be down around 10%. So when BTC was trading around 3’500-4’000 in April you would have bought a large amount of BTC’s. In this trading system, I did not consider position sizing but I certainly trade that way. I will provide those results in another article.

Look how nicely it stays out of many drawdowns. The drop from 14’500$ to 5’900$ it was completey out of the market. Same goes for the drop from 6’500$ to 3’200$. The market plummeted but the system is flat. On the other hand, the massive bull run from April on the model was fully invested and currently it would still sit in a long trade (it shows an exit here on June 30th but that exit was forced to be considered for my stats; it is actually still long since we had not taken out a 5-day low).
Dont let the simplicity fool you. I could give you highly sophisticated orderflow based trading models but it wouldnt do you any good. This is simple and anyone can follow it and in trading simplicity is king.

The best period to buy Bitcoin

0

When I am analyzing a new market to trade, I always start out with the fundamentals like, whats the average range (actually I look for average swing but thats for another article) per time period, average volume and its standard deviations or also which period of day is most volatile/active. Especially in 24hour markets like FX (or Crypto in this case) it can really help to split up the day in 3 sessions, the Asian-, European- and US session. Working in one of the top 5 wealth managers in the world, I can tell you that we certainly care which session we are in. FX books (not the ones you can read in ;-)) are handed over from our Asian guys to the Traders that are handling the Euopean session who in turn hand over their book to our guys in NY & Chicago. If we had to execute a large trade, we often waited for certain time periods (market closes/opens) when liquidity was greatest.

In this article I will not focus on liquidity but rather which time period is the best to buy Bitcoin and which is best to sell. I will use data from the 01.01.2018 until the 30th of May 2019. At the time of this writing this leaves only about one month for out-of-sample testing but the goal is not to create a full trading strategy but rather do some research. As time passes more true out-of-sample data will be available and also the period before 2018 can be used for further validation.
During the above period Bitcoin went from 13’850$ on the 01.01.2018 down to 3’215$ and since April 2019 it adavanced up to 12’215$. The buy-and-hold performance for this period was a loss of just over 1’600$. 

Which time of day is best to buy Bitcoin?
I had mentioned the 3 trading sessions above. Lets see how the performance would have been if one bought every day at a certain time and held the trade for 8 hours, a 1/3 day.
All times are Berlin times.
Best Time to buy BitcoinThe Table above shows the performance of buying 1 Bitcoin and holding it for 8 hours. Row 14 shows the highest profit of 5313$ during a period (01.01.2018 -30.05.2019) where Bitcoin buy-and-hold lost 1600$. Entrytime “0” means buying BTC at 00:00 Berlin time.

What can be seen is that the Asian session is clearly the worst performer while the later European and Early US sessions are better.
Lets break it down even more and look at hourly performance (buying and exiting the trade after 1 hour):
Best hour to buy bitcoin
Again the worst performer seems to be the hourly periods during the Asian session. Best are European Lunch times and US evenings.

Which day of week is best to buy Bitcoin?
Lets quickly look what day of week is best to buy Bitcoin.
0=sunday, 1= monday, …
What is very obvious is that friday and saturday were the only weekdays were it was profitable to buy bitcoin (entry was on 00:00 berlin time, exit 24hours later).

Conclusion:
So what now with all this info? There are multiple ways of using this. Seasonality like this study actually works well in other markets like commodity futures and equities. I have successfully used this as filter for some equity index futures models. In equities, what works quite well is to only go long towrwards the last few days of the month or the first few days i.e. from the 28th to the 2nd. A reason could be that many IRA accounts are buying at the end of month which creates demand. What also works well is trading before important news events like the FOMC or Futures/Options expiration days. In crypto, it seems that buying only on a friday or saturday could be a filter if one wants to reduce the exposure in the market.

Does the NYSE Net Volume influcence Equity Day Trading Models?

0

Similar to that post here, Im analyzing the impact of the NET volume traded on the NYSE and how it is impacting day trading models. The Net volume on the NYSE is essentially all the USD Volume that is flowing into stocks that trade higher minus all the USD Volume of stocks that are trading lower.

I am using the same dummy day trading strategy as previously as a benchmark. The rules again are:
If the Low is lower than the lowest low of the last 5 bars then BUY SPY
Profit Target 100$
Stop Loss 100$
Maximum time in trade = 2 hours
Close all open trades at the end of the regular session

Symbol: SPY on 5-min
Symbol 2: NYSE NET Volume (Symbol at IQFeed: DINN.Z)
Period: 01.08.2015 to 01.08.2017 leaving the rest of data for Out-of-Sample testing
Capital: 100’000$
Amount per trade: 100’000$
No commission or Slippage
Max Bars back: 200

In case you are wondering why the benchmark results are not exactly the same, the reason is I have a different max bars back setting.

The most important performance figures to me are Profit factor, ROA and Avg. Trade.
Profit factor: 1.07
ROA: 286%
Avg. Trade: 3.45$
Trades: 4371

1.) Buy SPY on a pullback (Benchmark Strategy) when the NYSE Net Volume makes a new x bar high (Switch in last column below = 1) or low (Switch = 2)

Variation 17 is the benchmark strategy without  the above filter. What can be immediately seen is that it is much better to buy the SPY when the NYSE net volume is making a higher high (Switch = 1 means buying on a higher high). Look on the last column to the right and see how many variations are above the benchmark strategy (I sorted it via the Profit Factor) with only 4 variations below it. It is also clear that you should not buy the SPY when the NYSE Net Volume is making a lower low. Basically it means it is much better to go long when there is more volume flowing into up stocks than down stocks and this makes intuitively sense.

2.) Buy SPY when the absolute NYSE Net Volume is above (Switch = 5) or below (Switch=6) x millions

Row 16 shows the performance when the NYSE Net Volume is > 0. Profit Factor, ROA and Avg. trade are all already higher then the benchmark strategy. What is also totally obvious is the fact that every single variation where NYSE Net Volume is below 0 is worse than the benchmark strategy and once Net volume gets below 7 million the performance is even negative. I really like the linear relationship showing the more negative the net volume the worse the performance. This could be a real edge but needs to be validated on out-of-sample data (see below).

3.) Buy SPY when the NYSE Net Volume makes a new high
I actually expected this to be more profitable than then benchmark too but its actually not the case. There had been 452 Trades with a total net Profit of only 705$. Profit Factor was 1.03, Avg. Trade $1.56 and ROA 34%. Pretty surprising this underperformance.

4.) Buy SPY when the NYSE Net Volume is above (Switch = 11) or below (Switch = 12) an x period bollinger band
Although it shows a similar behaviour that you are better of buying when the NYSE Net Volume is above its x period Bollinger band, only very few variations show an outperformance. Parameter 1 means the 5 period BB, 2=10 period, 3=15 etc.
Out-of-Sample Validation:
Lets check the 4 models above on out-of-sample data from 01.08.2017 to 01.02.2019.

The benchmark strategy performance for the out-of-sample period:
Net Profit: 7471$
Profit Factor 1.04
ROA: 120%
Avg. Trade $1.94
Trades 3846

1.) Out-of_sample: Buy SPY on a pullback (Benchmark Strategy) when the NYSE Net Volume makes a new x bar high (Switch in last column below = 1) or low (Switch = 2)
Variation 15 shows the benchmark strategy. Unfortunately it shows completely the opposite behaviour since now all variations perform better when the NYSE Net Volume makes new lows. This shows how important it is to do solid testing and not fool yourself. When I am developing fully automated models I do a lot more stringent testing but this would already be a model I wouldnt trade since it clearly failed out of sample.

2.) Out-of-Sample: Buy SPY when the absolute NYSE Net Volume is above (Switch = 5) or below (Switch=6) x millions

The best pattern from the in-sample test also help up out-of sample. Although it has less variations that outperform the benchmark, it certainly seems important to monitor the absolute value of the NYSE Net Volume. If it is above 15 million then it shows a very high edge although this only occured 28 times together with the Spy pullback “dummy pattern”. What I really like is the linerarity of this pattern. Net volume of 15 million is better than 14 million is better than 13 million and so on. All down to 7 million seems a good edge/filter for long trades.

3.) Out of sample: Buy SPY when the NYSE Net Volume makes a new high
The results were mixed and not really better. Same as in-sample.

4.) Out of sample: Buy SPY when the NYSE Net Volume is above (Switch = 11) or below (Switch = 12) an x period bollinger band
90% of all variations show an outperformance vs the Benchmark strategy with higher profit factors and higher avg. Trade. However there is no linearity and i prefer if the opposite signal is clearly negtive. Mixed bag here, if you ask me.

Overall conclusion:
Although I am a bit surprised that Pattern 1 completely reversed, we still found something that can be of incredible use. Monitoring pattern 2 and where the NYSE Absolute volume trades seems important. When NYSE Net Volume gets above 7-10 million this is very supportive of bullish strategies and should be monitored.

Improving your Day Trading with the NYSE TICK

1

I am analysing how the NYSE TICK indicator can be used to improve certain day trading strategies. The NYSE TICK measures all equities on the NYSE and whether the last trade occured on an uptick or downtick. Its a good measure of the overall stock market action and what I am hoping to find is a usecase as a filter for day trades.

I will use the SPY ETF as a proxy for the ES (thats the S&P500 futures). I do this for multiple reasons. First, Im always concerned about over optimization and using a releated market limits that to a small extend. Second, the NYSE TICK measures the Equity markets which are in a different timezone than the futures in Chicago and the SPY trades in NY too, so that makes it easier (a software thing).

Symbol: SPY on 5-min
Symbol 2: NYSE TICK on 5 min (JTYT.Z on Iqfeed)
Symbol 3: NYSE TICK on daily time frame
Period: 01.08.2015 to 01.08.2017 leaving the rest of data for Out-of-Sample testing
Capital: 100’000$
Amount per trade: 100’000$
No commission or Slippage

Benchmark Strategy Rules:

Since my daytrading methods for equity index futures are mainly mean reversion (counter trend or fading moves) I will create a similar dummy strategy to act as a benchmark and to see how it could be improved with the TICK.

If the Low is lower than the lowest low of the last 5 bars then BUY SPY
Profit Target 100$
Stop Loss 100$
Maximum time in trade = 2 hours
Close all open trades at the end of the regular session
The benchmark strategy yields surprising positive results but it is untradeable as the profit per trade is only 3.42$ and there are no slippage and commission considerations. However this strategy was not intended to be traded but rather serve as a benchmark which mimmicks my personal trading of Index futures to a certain extend.

Testing different patterns with the NYSE TICK

Buying the SPY on a pullback if the TICK made a new highest high or lowest low
What if i only buy once my benchmark strategy condition is true AND the TICK makes an x bar highest high or lowest low?

Variation 7 shows the Benchmark strategy. Switch (last column) shows that TICK made a highest high (if Switch = 1) or lowest low (if Switch = 2). So for example row 1 shows the Benchmark strategy (buy SPY if low is lower than the last 5 bars) only when the TICK made the highest High of the last 28 bars. This only happend 28 times which is too few to be relevant.
What can be seen that it seems beneficial to only buy the SPY when the TICK made a short term (not excessive) low of 4-10 bars (Rows 4-6 all have higher Avg. Trade, ROA and Profit Factors although none are significantly better). What is also obvious is that it is not a good idea to buy the SPY when the TICK made some significant new lows i.e. if the TICK made a new lowest low for the previous 28-57 bars or more (not shown here).
Buying when the TICK made significant new highs seems better but there are only a few trades which make results more unreliable. Also consider this is a pure optimization on an in-sample basis and is likely significantly curve fitted. If these findings have any merit will have to be verified on out-of-sample (unseen) data.

Possible (in-sample) Conclusion: Do not buy the SPY when the NYSE TICK is very bearish, making significant new lows. Only buy if the TICK made a short term low (2-10 5min bars) or if the TICK made a significant High (>25 5min bars High).

Buying the SPY on a pullback if the TICK CLOSED on a new highest/lowest CLOSE
That test didnt bring any more relevant results and in my opinion that makes sense. The TICK measures every stock in the NYSE and if its currently up ticking or down ticking and is calculated at a certain time interval (every 6 seconds on the NYSE but certain data vendors might have different intervals). I wouldnt see why it mattered where it “closed” that 6 seconds and highs and lows make a lot more sense to me.

Buying the SPY on a pullback if the high of the TICK is above/below a level
Lets test the benchmark strategy and its performance if the High of TICK is above a certain level to buy and also if the low is below a certain level to buy.
Row 1 shows if you buy the SPY on a pullback when the High of the TICK is above +100. Row 2 the TICK would be above +200 etc. And from Row 16 it is the counter trend entry, so in order to buy the low of the TICK would need to be below -100…
What is evident here is that it is beneficial to buy the SPY if the high of the TICK is above a certain level. If the low of the TICK is below a certain threshold, it clearly underperforms.
Quite an improvement if the SPY is bought when the TICK High is above 100.

Buying the SPY on a pullback if the TICK makes a new intraday high 
This could be classified as a divergence where the SPY makes a 5 bar low while the TICK makes a new intraday high.
The profit factor of 1.33 is the highest from all variations so far (if at least 100 trades were generated) and also the Avg. Trade with >14$ is the highest. However this only occured on about 200 trades.

Buying the SPY if the high of TICK trades above/below the Bollinger Band
No interesting results. None improved the benchmark model

Summary from in-sample tests:
It seems the TICK has a usecase to improve day trading models for Equity Index instruments. The main findings so far were:

  1. it is better to buy the SPY on a pullback if the TICK makes a short-term low of 4-10 5min bars
  2. do not Buy the SPY if the TICK makes a significant low >27 bars
  3. buy the SPY if the TICK made a significant new high > 25 bars
  4. buy the SPY if the high of TICK is > +100 or more
  5. do not buy the SPY if the low of TICK is below -100 or less
  6. buy the SPY if the TICK makes a new intraday high

Lets now check if these findings hold up also in Out-of-sample tests. I am now changing to unseen data from 01.08.2017 – 31.01.2019 for these models.

The Benchmark Strategy Return is:
$8376 Profit
Return on Account 134% (that is basically the Return vs. Drawdown)
Profit Factor 1.04
Avg. Trade 2.16$
Trades 3876

Out-of-Sample verification of above findings:

  1. Is it better to buy the SPY on a pullback if the TICK makes a short-term low of 4-10 (5min) bars:
    The pattern holds up as all variations have a higher ROA, higher Avg. Trade and better Profit Fator.
  2. Do not Buy the SPY if the TICK makes a significant low >27 bar:
    This behavior also holds up as all ROA, Profit Factor and Avg. Trade are worse than the Benchmark Strategy.
  3. Buy the SPY if the TICK made a significant new high > 25 bars
    This definitely is also confirmed on an out-of-sample basis as the ROA is triple that of the Benchmark with a high Profit Factor, Avg Trade that is 12 times higher and also a much higher win rate.
  4. Buy the SPY if the high of TICK is > +100 or more
    This behaviour also holds up however the value that the high of TICK has to reach needs to be higher. From the results below it can be seen that the higher the TICK Reading the better the performance figures. A value of 3 in the second column from the right means the high of TICK needs to be above 300 (4=400,…). Readings of 100 and 200 are not yet good enough but from 300 on all factors are better.
  5. Do not buy the SPY if the low of TICK is below -100 or less
    This behavior also keeps repeating as the lower the Low of the TICK gets the worse it is to buy the SPY.
  6. Buy the SPY if the TICK makes a new intraday high
    this seems to be the only varation where one can not clearly say it held up. Although the avg. Trade of 2.78$ is better than the benchmark with 2.16$ and the Profit Factor of 1.06 is also slightly better, the ROA with 26.79 is worse.$

Overall conclusion:
Almost all of the identified patterns improved the benchmark strategy, sometimes significantly. What was surprising was that 5 of 6 patterns clearly held up out-of-sample and the 6th was kind of a mixed bag. The TICK seems to be a good filter for taking daytrades in the SPY and should perform similarly in the e-mini S&P500 so it definitely seems worth it to keep an eye on it when trading Index products on a day trade time frame.

HODL vs Trading: Is it Better to Trade Bitcoin?

0

I did a study on Bitcoin and was looking for a way to reduce price risk, yet still fully participate in the lucrative gains that Bitcoin has seen in the past. I essentially wanted to take the guess work out when to be invested and when to take profits, all considering that I am a long term believer of blockchain and my desire to participate in this trend.

Benchmark performance: HODLing (Buy-n-Hold) Bitcoin
I chose to look at two years of data from 01.01.2016 to 31.12.2017 from Bitfinex. I ignored commissions and slippage in this analysis and bought 1 bitcoin when the price was about 425 USD.
So HODLing (buy-n-hold) bitcoin would have produced stellar gains of over 648% annually in those two years. However, the drawdown would have been 9173 USD of your portfolio value. Not many people can easily stomach giving back almost half of their gains and people who have been in the cryptospace long enough, know that there have been times with much larger losses.

But what if there is a way of reducing those large losses yet still being able to participate in the upside gains?
I created a simple strategy (I am a fan of simplicity especially in Algorithmic Trading) that uses the following rules:

Buy BTC when it trades (1 Tick) above the 10 day High.
Sell BTC (close long trade and be flat) when it trades (1Tick) below the 5 day Low.

That’s it. No other rules. Lets look at the performance:
What we can see is that the return is slightly higher with 11980 USD gained but the drawdown of 4412$ is cut in half, a major improvement.
The input parameters for entering (10 day high) and exiting (5day low) the bitcoin position were randomly chosen. Lets check if those parameters can be improved. I tested in 5 day increments (so entering on a 5 day high, 10 day high, 15 day high etc.):
Looking at the test results from above one can see that entering at a 10 day high and exiting when bitcoin trades below a 5 day low is not ideal as more profits can be generated if one enters and exits at a 10 day high/low. However this is where it gets tricky. My original idea was to enter when bitcoin makes a new 10-day high and get out of the position when it trades below the 5 day low. Switching now to a new rule set of 10/10 becasue of a backtest could lead to overoptimization (also known as curve fitting, overfitting) and in systematic trading you want to avoid overfitting like the plague. I will describe the challenges of curve fitting and ways to deal with it in another article in the future.
So we have to be careful not to curvefit the model, yet there are a few things I like about the other parameters. The main benefit I see would be a reduction in trades. This is a longer term investing/trading strategy for bitcoin and I dont want to be too active as it increases trading costs (commission and slippage) and would require having the bitcoins more often at the trading Exchange instead of a cold storage like a Trezor, Ledger or paperwallet. And Bitcoin or Altcoins should always be stored in a cold storage wallet if possible. For the sake of this analysis I will leave the trading methodology as is (enter on 10 day high, exit on 5 day low). Since the CME & CBOT offer Bitcoin futures we can actually leave the bitcoins in cold storage and hedge the risks via futures. I will describe how this works in a future arcticle. If you want to use this trading model but with parameters where you dont have to trade that frequently I would use parameters like 10 (exitlevel) & 20 (entry level) so one would have very similar results (over 12000$ profit with 4954$ of drawdown) but with a lot less trades.

Lets add more data and see if the original method still outperforms HODLing Bitcoin.
I added data from November 2013 until today the 25th of February 2018.
The benchmark again is HODLing where one buys 1 BTC on the 16th of November 2013 (thats the first day of somewhat clean data from Bitfinex) for 420$ and holds it until yesterday (24.02.2018).
So buy and hold Bitcoin from November 2013 until today yielded 9723 $ with a hefty drawdown of 13891$.

The simple Bitcoin trading strategy where you buy when Bitcoin makes a new 10 day high and sell when it breaks below the 5 day low significantly outperformed HODLing on an absolute and especially on a risk adjusted basis as can be seen below:

The periods from 2013 to 2016 and 2018 can be seen as a so-called out-of-sample test where we used the same parameters as before but on “unknown” data.

Below is a screenshot of the most recent trades and as of 25.02.2018 the trading model (5 & 10) is out of the market:

Lets also peak at the other entry/exit parameters over the whole period from November 2013 until today (February 2018) and how they would have performed:
From the 25 possible parameter settings, 22 of them outperform buy and hold on a risk adjusted basis.
The 10 & 20 model mentioned earlier also significantly outperformed the benchmark.

Conclusion:
While HODLing is viable approach to investing in Bitcoin, using some form of exit methodology might be easier to handle psychologically as drawdowns can be reduced and even profits could be increased. Although this method seems to be robust (it has only 1 entry and exit rule and performed well on an out-of-sample basis) there cannot be any guarantees that this will still be the case in the future. The model will actually lose (underperform) when bitcoin trades sidewards constantly as multiple stop outs could occur.
What makes me believe that this model will keep performing well is the fact that Bitcoin experiences extremely strong price moves, something I dont expect to change dramatically in the near future. I like the simplicity of the trading model and if one does not have a futures trading account to hedge risk (i.e. go short futures when an exit signal appears while keep holding the “spot” bitcoins in a paperwallet) then he/she could look at using 10/20 parameter setting (exit at 10 day low, enter at 20 day high). One thing seems likely, if Bitcoin really does reach heights of 50’000$-100’000$ or higher in the next few years then this model should keep doing well. Until then, best of luck

***Trading Futures-, Options-, Equities and retail off-exchange foreign currency or crypto currency transactions involves substantial risk of loss and is not suitable for all investors.You may lose all or more of your initial investment. Opinions, market data, and recommendations are subject to change at any time and should not be construed as an offer or a recommendation to buy or sell a security nor is it to be construed as investment advice.

My Approach For This Trading Exercise

0

As a discretionary trader I highly believe in the edge of beeing able to read context. After all, I think thats the main advantage we have over algorithmic trading strategies. However, for this exercise this also provides some challenges, mainly how do I trade an approach based on market context mechanically? After some thinking I decided that I would trade like this (for this exercise):

Pre-market I would mark up my charts with important supply and demand levels, where I expect a reaction. Once trading at one of those levels, I would use a mechanical entry trigger to get into the trade. I dont expect getting into trades at each level but only where I get an entry trigger. After entering the trade, my trade management is also completely mechanical with fixed targets and stops.
I trade 3 markets at the same time which are the ES (e-mini S&P500), NQ (Nasdaq100) and the YM (mini-DOW). In the ES i use a fixed 2pts target and 2 pts stop. In the NQ its 4pts target and stop and in the YM its 15pts for both target and stop loss.
I am aware of the fact that my trade management is sub par as commissions will cause my avg. winner to be slightly smaller then my avg. loser, however for the purpose of this exercise I dont care. I also have the strong feeling that my winners would have much more potential but I havent quantified that yet. My backtesting which I did on a bar by bar basis, shows a positive expectancy, one that almost looks too good to be honest but the human mind can be evil. I realize that subconciusly I could have cheated (f.e. seeing a good setup but it was a tick away from the supply/demand area and I still took the trade in my testing, or news events like NFP, FOMC caused some troubles etc.) because my mind is so focused on having an edge, that this could lead to cheating myself without counciously knowing it.
However, all of that doesnt matter too much for that exercise, as I wont be trading like that in the long run anyway. The only part I will keep are my S/D levels to do business at, which is the basis of this approach also and the goal for this trading exercise is to learn to execute my plan without any errors.

Mechanical trading approach

Here is a screenshot of my backtesting. Its based on trading 3 markets with 2 contracts each, but I will only trade 1 contract. Although the statistics look good, there is no commission and slippage included which could have drastic effects on the performance. While the expectancy per contract at USD 31.60 is certainly high enough to cover commissions, slippage could have a pretty bad effect. For example, I could enter 2-3 ticks worse because of slippage but my target in the YM would still be 15 pts and while my testing could have potentially resulted in a winning trade, because of slippage this might be a loser.
Never-the-less, Im still confident of trading like this for the limited sample size of 25 trades. Like I mentioned before, if all 25 trades would be losers it would cost me 2500USD, a risk I have fully accepted.

“Thinking In Probabilities!”

0

Trading can be simple, but its not easy (to say the least)! At the very core, you “only” need two things:

1.) A trading approach with an edge;

2.) The skills to flawlessly execute that edge;

Thats it! Not much else is needed. However, anyone who actively trades knows how tough trading can be and people spend years (or their entire life) searching for their edge (most are blinded by the search for the holy grail)!
In contrast, people spend very little time and energy to develop that so important second part, the skills to FLAWLESSLY execute an edge. Thinking about myself, I believe I fall exactly in the same category as most traders. I have spent the absolute majority of my time and energy on part 1, trying to find an edge. The interesting point is, even when I was making money trading a certain approach, I quickly put it aside again (or got distracted), thinking that there must be “a better way” of trading the markets. Looking back I can clearly see that I didnt know my edge because I didnt trade a certain approach long enough to be able to even quantify it…
Although I certainly could have used my time and energy much more deliberately, I dont regret looking at various approaches. After all, I have a pretty good picture of what is out there and I feel  I have learnt a lot. Most importantly I have found a way of looking at the markets that makes total sense to me. To me the markets are an auction wether you look at it from a very macro point-of-view or you drill down to the micro levels of looking at orderflow etc. For me its all about supply and demand so I dont really care about what current moon cycle we are in or what Ganns Square of Nine is… Im not saying those things dont work, they just dont work for me because although I understand the mathematial equation behind it, I dont understand why that should have any predictive power. There are a million ways of approaching the market and I will focus on one that make sense to me!

Now going back to the subject of this post… I am a descretionary trader and I think the biggest challenge will be to remove the randomness from my trading. Please dont be misled by my words here. With randomness I dont mean that I have some sort of “shoot from the hip” approach, where I randomly enter and exit a trade. I think many people, especially those who trade algorithmic trading systems, have a wrong picture of discretionary traders. They think a descretionary trader just randomly enters and exits the market but I doubt anyone could be successful at that. Not even those who have advanced to an intuitive level of trading which is in my opinion the highest level of trading expertise one can reach, trade purely on intuition. I believe they have a pre-market plan and because of their vast amount of deliberate trading experience, they are quickly able to “get in the zone”, a state where they can extract profits from the market almost at will!
My personal challenges are that although I always (hmm, maybe not always, I should write “mostly”) :-)) have a well thought out plan, I still have too many factors that are not rigid. I know where I want to get in (the trade location) and what my risk will be but my trade management needs improvement and so does my timing for an entry. If some parts of my trading plan are not totally rigid, then there is a chance that part of my trading is random, something that I must avoid. The main reason I must avoid that is because it will be almost impossible to improve and get better. After all, even when journaling every trade, I will not know what caused an undesirable result. Was it my trade management? My timing? Did I read context wrong? An impossible task if my trading rules are not more clear cut.

I am committed  to change that once and for all even if that means to go way back and trade a very simplistic and mechanical approach where the only subjective decision making is done pre-market, when I dont have a position and thus can not (not very much) be influenced.
Mark Douglas, my favorite trading psychology resource, calls this execise “The 5 Mile Rule”. It is well explained in his superb book “Trading In The Zone” and his DVD “How to Think Like a Professional Trader”, both highly recommended.
What you do is quite simple: You take a completely rigid and tested trading approach, one where there is hardly any subjective decision making to be done and trade that in limited sample sizes of 25to 30 trades, no matter what. The purpose of this excercise is not necessarely to make money, but to learn the skill of flawless execution! My focus will therefore be on just executing my plan and fully accepting the risk. Speaking about risk, the maximum risk per trade will be 100USD in the ES, 75USD in the YM and 80USD in the NQ, trading only one contract. So hypothetically speaking, if I would only get trades in the ES (highest risk because of the 2pts/100USD stop) and all of those would be losers, I would lose 2500USD, a relatively cheap price for truly learning to think in probabilities.

Social Media

6FollowersFollow