I tested the 25 best strategies over 2023's data - This is what I found !! (Patreon)

Hi there and welcome to this video. The start of a new year finally  gave me the opportunity to execute an idea that I had for a long time.

In this video I will show you the tests of the Top 25 best trading  strategies and their theoretical performance if you would use them in  2023.

I will also compare these strategies results with a bitcoin investment baseline over 2023.

This way I can compare a simple medium long term investment in  Bitcoin against the performance of trading strategies and find out which  method creates more alpha.

And believe me, in this video you will see some very surprising  results. So watch until the end to find out what my end conclusion is  and what to take into consideration for the next bull cycle.

Introduction

In the last year I regularly  received the question why I did not include the data of 2023 when I was  testing out the performance of algorithmic trading strategies with  backtests.

The first reason that I want to compare each and every algorithm over  the same datasets so that I have an honest comparison over all the  strategies.

It wouldn't be fair to add extra data every time I test a new  strategy, because that could influence the end result in a positive or  negative way. Depending on the direction of the market at that time.

And remember that I will keep on using the original datasets in the  future when I test new strategies performances. Just to keep these  comparisons honest.

Please keep that in mind when you are watching future video's.

Now the second reason I wanted to wait this long is the result you will see in this video.

I deliberately waited until the end of 2023 because I wanted to  analyze the performance of the top 25 trading strategies over the  complete year.

This way I think I have an even better idea of the performance of each strategy in comparison with the old data.

Just as with machine learning as explained in my video on FreqAI you  can use a piece of a dataset to train a model and a second piece of a  dataset to test.

Now in my case I could train, or rather initially test the algos with  data before the start of 2023. But the dataset to test over 2023 was  not complete until the first of januari 24.

So that's why I had to wait until this moment to finally be able to do these tests.

The setup

Now what about the setup for this test.

Let's begin with the strategy league and the algorithms that I will use in this video.

At the end of 2023 I had the following league table, including the Obelisk Ichimoku slow results of my last video.

I have put aside all these strategies files including their optimization settings whenever I optimized them.

I will test each and every of these strategies in the order of  position, starting with position 25, the optimized "Smart money  strategy". And I will then work my way back to the first position. The  "Obelisk Ichimoku Zema".

However I will present the tested strategies in the order of their new position.

Making a countdown to the strategy that has the best performance over the data of 2023.

In theory that strategy would have the highest probability to make money with trading during last year's rally.

The timerange and datasource

Now what about the timerange and the origin of the data?

Well, the data that I have downloaded comes from Binance and I downloaded all that data on januari first 2024.

The timerange of this data and thus the backtest is 20230101 - 20240101.

I downloaded this data with a similar command like this one you see here:

So exactly a year's worth of trading data.

The Freqtrade config has all information about which exchange to use for data downloading and the pairs to download data for.

Pairs data

And speaking of pairs. As you  might already know I always use 50 pairs for testing out a strategy with  a maximum of 10 trades at the same time.

I could have kept this maximum open trades at the default of 3 but I  changed it to 10. And that was just an arbitrary, random choice at that  time. From that moment on I just kept this setting.

I thought more about the choice of pairs to use for these tests. I  wanted to  use pairs that provided enough liquidity  so that I was able  to step into and out of the market easier. Eventually the pairs were  chosen from the coinmarketcap snapshot of januari second 2022 .

Now not all pairs you seen from this page were available on Binance,  so I kept adding available pairs from this list until I got 50 pairs in  total.

Spot pairs

For spot trading I used the following list of pairs:

Futures pairs

And for futures trading I used the following list of pairs, since not all spot pairs are available as futures pairs.

I also have a remark on the availability of some of these futures  pairs because during the period when I was determining this list and now  some pairs weren't available anymore on Binance. Or at least I could  not get data for them anymore so I had to comment out some of them. And  so these will not be used in the tests unfortunately.

By the way, all this information, including data, configs, settings  and more are all available for my patrons on my Patreon page. This way I  try to be as transparant as can be. Everybody can have a chance to  reproduce and check my backtest outcomes on their own machine. And if  there are any questions or problems with this I am happy to help to find  a solution for my Patrons.

Timeframes

Using this backtest timerange and  these pairs I configured, I backtest over multiple timeframes. The  reason for this is that I noticed that sometimes a strategy that was  build for the 5 minute timeframe performed much better on, let's say,  the 1 hour time frame. And vice versa.

These timeframes I eventually chose to use aree the 1 day, 4 hour, 1 hour, 30 minute, 15 minute and 5 minute timeframe.

I could test over other timeframes as well like weekly, 12 hour, 1  minute and so on. But I chose these other ones to balance out the  timeframes for long and medium term swing trading and shorter timeframe  scalp trading.

Again this is an arbitrary personal choice and gives me a good view on which timeframe an algorithm performs best.

There is one other remark I have to make and that is the difference  of timeframe that ended up in the strategy league and the timeframe that  worked best in 2023. There will be strategies that had best results on  the 30 minute timeframe for example and in this test the 1 day timeframe  performs better. I did this deliberately to demonstrate that the  timeframe is an important factor for the profitability of an algorithm.

Some youtubers or other articles tell you that their strategy only  works best on a specific timeframe. I wanted to demonstrate that you can  easily use someones trading idea and use it on other timeframe as well.

Scoring

After the backtest of each algorithm the  backtesting engine of the Freqtrade bot provides all the performance  information for that strategy and timeframe.

To compare these results between timeframes for the same strategies  and the performances between different strategies, I created a scoring  mechanism with which I can make performance comparisons.

Again the way I score certain results are also configured to my personal risk tolerance.

If you want more information about this performance scoring, I have an extensive Patreon blog post that explains it all.

I will use the same scoring for this video as I did in the other  videos. And the results of the backtests and their corresponding end  scores will determine the position for this video on the temporary  strategy league.

Start testing

So, now that I provided you  with some background information about these backtests, lets begin with  the results of the strategy that came last on the Top 25 list.

Do not expect an in depth explanation of these strategies. For that you can visit my youtube channels playlist and watch the corresponding video about that strategy.

Using a baseline

Let's first watch the  total markecap of crypto to see how the general price of Crypto overall  moved in 2023. If we use this as some sort of a baseline for the market  movement, we can then compare this with the performance of the  individual strategies.

As you can see, from the beginning of 2023 the total market cap had a beautifull rise of over 131 percent.

Ofcourse there were some dips that were great buying opportunities,  but overall this has a classic higher high, higher low structure. With a  small consolidation between august and november.

It's no coincidence that this total marketcap closely follows the  price of Bitcoin. Since Bitcoin is the king of crypto and still has the  highest marketcap.  So let's use this information to create a baseline to which we can  compare the performance of the strategies I will show later.

Say that we start 2023 with 1000 dollars for trading and 1000 dollars for investing in Bitcoin.

So say that I bought and held 1000 dollars worth of Bitcoin on  january first at a price of 16.499 Dollar. Selling that Bitcoin amount  exactly one year later, at a price of around 44180 dollar, gives us a  profit of around 170 percent.

And when I mutliply 1000 times 170 percent, that means that in the  end we have a profit of 1700 dollar profit. Which is not too shabby at  all.

In other words my end balance on my Crypto exchange should be the  initial balance plus the profits. Which is around 2700 dollars.

If I make this 170 percent gain a baseline to test the profitability  of the strategies against. Than I can determine if trading could be  profitable against investing. And also which strategies I better could  have ignored in 2023?

Let's begin!

Let's now concentrate on the  strategies I have tested the last couple of days. Could we have gained  more or less than 170 percent profit if we used one of the Top25 trading  strategies? And what would the difference be?

And also how would the profile of the account balance look like over  2023. Profits right from the start and no looking back, or a nail biting  psychological nightmare..?

Let's find this out next...

25. Obelisk_Ichimoku_Slow_v1_3

The first strategy that I will show you and that ended up on the 25th spot is the Obelisk Ichimoku slow.

In the strategy league where a larger dataset was used, this algorithm takes up the 15th position.

However if we were to use this strategy in 2023, this would actually  provide us with a loss of 15 percent. Ending our balance at 848 dollar.

The 30 minute timeframe would still be the best timeframe to use this strategy on. But no matter how you look at it.

This first result directly makes the important statement that  backtests over data in the past, and the results and conclusions that  can be done on them, do not provide any guarantees that the same  algorithm makes profits in the future as well.

This is exactly the reason why I regularly give warnings on the  results I present. They are hypothetical results that can, and listen  good, CAN make money in the future. But it is all a matter of  probabilities and no absolutes.

So lets take a quick look at the drawdown of this algo over the last year, which has an average of 28 percent.

And the average profit and loss, that gets positive around november  with a couple of weeks that are in the green in october. But not enough  profits to make the end balance of the account positive.

So this strategy ends up as the last one in the test of the Top 25 over 2023.

Let's forget this one quickly and continue with the strategy that ends up as number 24 on the list.

24. skzft

Now this strategy was actually an  aglorithm that I created as a test for another algo that will be  presented later. It performed surprisingly well and I left the results  in the overall league, but I cannot remember I ever made a post or video  about it.

Not that it matters much because this one also did not perform very well over the data of 2023.

It did make some profits so you did not lose all your money. But  since we have established a baseline with the Bitcoin price we can  conclude that this strategy also did not perform great. Only a meager  1.5 percent profit.

Funny enough this algorithm had the 24th spot on the original league  and in this test it also has the 24th spot. So no change there.

Also know that this algo was optimized against historic data as well  and this optimization did not prove to be the gamechanger for 2023's  performance.

Finally the win/lose chart for this strategy does not give me much confidence over it's performance.

Maybe one might could cope with the losses in the beginning of the  year since it had some good winning trades. But after a while you would  see your earnings dry up and turn into losses.

And I can imagine that after a while you would stop with this strategy because you only see losses and try another one instead.

Probably somewhere at the highest point of the drawdown.

So with this knowledge and the 24th spot known, let's move ahead with the number 23 strategy on the list.

23. Stray Kids Spot and 22. Stray Kids futures strategies

And  these are the official brothers of the strategy I just presented. The  Stray Kids strategies on respectively spot and futures trading.

The spot version has the following results:

And the futures strategy these results:

The interesting part here is that there's not much of a difference in  gained balance between these two versions of the same strategy. Even  though one only does long trades in spot markets and the other version  does long and short with futures positions.

Remember that I told you that the pairs of both these versions are  not exactly the same as well. It could be that more Futures pairs would  improve the performance of that version significantly.

Looking at the weekly historic performance of the strategies makes this more clear.

But overall, both these versions still perform very bad over the year  and I think more detailed investigations are needed why these  performances lag behind.

The spot version loses straight from the start and only a couple of  lucky trades saved your trading balance at the end of the year.

Which is lucky but at the same time is actually something that you do not want to see.

A more detailed chart shows that, at the end of the year, the  strategy had some lucky trades with AVAX - ICP and KSM and that made  this algorithm still end up in the plus.

The same goes for the future oriented strategy version.

Although not as bad performing as the spot version, it still would  have tested your faith in the algorithm over the most part of the year.  And I do not beleive that adding a few altcoins into this equation would  make that much of a difference.

If I would see the market rise and rise but the strategies  performance stayed behind, I do not think I would keep the algo and hope  for the best. Just as with the previous result I showed.

Then I would count on luck and not op probabilities. And that is not the way I think about trading.

In this situation the end balance was saved by some good trades on  SOL, ICP, NEAR and some other coins. But again having a few lucky trades  is not the point of having a well thought out algo.

In my opinion it should provide a feeling of consistency and preferably a trading balance that grows steadily but surely.

But that is clearly not the case here with these two strategies, that  originally are placed at 10th and 16th spot in the official league.

So let's continue with the next trading strategy on the 21th spot.

21. keltner channel strategy

And  that strategy is the Keltner & RSI strategy. In the original video I  also optimized the parameter settings for the indicators used. And for  this excersise I also tested with these optimized parameters.

But again we see the same kind of graph appearing. As if a pattern emerges for these strategies I've shown so far.

From march until november there were more weeks that were in the red than in the green.

And the average profit over these weeks became positive from around oktober. Eerily similar as with the other algos this far.

Now again the drawdown was highest around the third quarter of 2023.  Which is contradictory of the crypto market development we saw in the  beginning of this post. And again this would put your faith to the  test.

Also the overall performance of this strategy is not quite mentionworthy

In this case the end result is again greatly influenced by a couple of good trades at the end of 2023.

And I'm starting to think at this point that, with these strategy  types, there should be an additional indicator or an additional signal  used, that would detect a point where you enable or disable these  strategies.

To prevent drawdowns and improve the probability of better performance and profitability here.

But maybe if we come closer to the top 10 or top 5 strategies, I have  more information to make a better hypothesis on what happens here.

At this moment we only have seen 5 performance results and are not even in the top 20 so anything can still happen.

20.PowerX strategy

The Power X strategy was originally placed at the 11th spot of the original league.

So 20th place does not give much confidence about the possible performance over 2023.

However, this strategy starts 2023 very positive with some nice positive weeks.

But after a while it loses it's edge and the account balance drops into negative territory.

Which results in a total drawdown of over 45% from it's peak. With an average drawdown of 26%.

Only around October does this strategy start to have positive weeks again, resulting in a positive end balance.

And again there is that same time period where this medium term  strategy starts to attain it's edge in the market again. Just like the  earlier strategies.

I will keep a close eye on this period when I am starting to analyze this on Tradingview in my spare time.

Perhaps there is a very obvious explanation that helps us understand this behavior.

However, the reality check is real and up until now there are no real survivors when it comes to the scores so far.

In my original league, scores of 125 and lower gives an algorithm a place in the lower 25% of the list.

And we are not even higher than 120 points at the moment!

So let's quickly move on with number 19.

Maybe this time we get some hope that things will turn around...

19. kijun cross only strategy with futures trading

The  Ichimoku Kinko Hyo is one of my favorite trading indicators and the  Kijun cross only strategy here is one of the variants that makes use of  this indicator.

This strategy was originally placed at the 23th spot on the league  but in 2023 it sits at the 19th spot. So let's quickly see if this algo  is worth our time...

But if I dive deeper into the portfolio balance over the year , then this view looks bleak and disappointing.

This strategy would put your patience to the test the whole year and  again only has some have a lucky shots at the end of the year.

This also means that the performance indicators of the strategy end  up positive. But this is a smoke screen on the overall performance  profile of the strategy.

It already started at the beginning of the year with a negative win  ration that remained the whole year. The winrate was 30% over this year,  which is the same as the original initial test of this strategy only  this time it is highly probable that the 'let your winners run'  principle could not be attained.

After the initial drawdown, small wins and losses kept the balance  relatively stable between minus 200 and 300 dollars. Except at the end  of the year. Then the true strenght of this strategy came into fruition  and it could detect trades that finally gave positive results.

Another aspect of the relative bad performance of this strategy is that only 16 percent of the pairs responded well.

So there should also be some further investigation why all the other pairs did not trade with this easy strategy.

I cannot imagine that a coin like Chainlink or Litecoin never had a  kijun crossover over the whole year. Even if it proved to be a losing  trade.

But let me continue because we still have 18 strategies to go.

So let's quickly move ahead.

18.The Sma Rsi strategy

The Sma  RSI strategy is yet another algorithm that scores reasonable at the end,  but has a suboptimal account balance profile over 2023.

During the year it suffers from more losing weeks than winning weeks  and its only around October that it once again starts to gain momentum  and ends up in the black numbers again.

You had to endure a maximum drawdown of 25 percent over the year before finally see this strategy's potential.

And as I said earlier. Just as with the other strategies that were  mentioned so far, this strategy also has a downwards portfolio balance  before finally starting to rise again around October.

The overall plot from the freqtrade bot shows us also the profile of  the strategies performance over 2023. And together with the score I give  to all these factors makes this algo perform better than its original  position in the strategy league, which was on the last spot. So in this  case the strategy performed better than the ones I previously mentioned.

17. Nostalgia For Infinity X

The next algorithm is a pretty popular one because the videos I made of its performance has one of the highest views.

It is the Nostalgia for Infinity algorithm with a whole bunch of buy  and sell conditions. And when I tested this algo out I expected it to,  at least perform as one of the best, since it has buy and sell  conditions for almost every situation...

But what surprised me the most was it's amount of trades over the  whole of 2023. Only 63 trades om the 5 minute timeframe, whereas the  PowerX strategy on the 4 hour timeframe has over 1300. Of which I think  is the other end of the scale with too much trades on a high timeframe.

Again, taking the weekly performance chart here I can discover a  little bit more about the way the algorithm responds to the 2023 market  circumstances.

In the beginning until around the beginning of may this algo had  positive weeks. But during may and june, where the overall market  circumstances were not so great. The Bitcoin price had a major  resistance at the 30.000 dollar level and the price had a pullback to  around 25.000 dollar. During that pullback this algo had some bad  performing weeks that caused the drawdown to rise to it's highest level  at 8 percent.

Which in comparison to the other strategies is not much. But it still caused a negative balance on the trading account.

The major factor why this gets the 18th spot is a combination of a  high win percencate, low drawdown good profit factor and the way all  crypto trading pairs respond well to this algorithm.

I still find it strange that this very complex strategy has this low  trade amount on the 5 minute timeframe and it would probably take a lot  of time to find out exactly why or why this not happens.

However, since this is the first algorithm that has a somewhat flat  profit performance curve and has no high drawdown, I cannot complain  that much. It is the first one that you probably still would keep active  over the year, although it does have a high opportunity cost in  relation to the baseline Bitcoin investment.

Now I have to point out that this does not mean that all the  strategies so far are worthless. It only means that these specific  algo's just don't fit the overall market situation that happened in  2023.

Over a longer backtesting they  would probably be profitable because  overall they scored well before 2023. It just happens to be that the  last bear market and the rise in 2023 do not fit these specific  strategies.

And that they could not show their edge under these market circumstances.

16.The Alligator Strategy

The next strategy that I have in the original Top25 and which scored the 16th spot is the Alligator strategy.

The end balance from this strategy is more than 1500 dollars.

On the first hand, and again I might tell you, the numbers at the end of the backtest do not appear to be really frightening.

And the Calmar, Sortino and Sharpe ratio's here show that as well.

But as you can see from the winrate and the drawdown, together with what we know from similar types of strategies.

You might suspect to see a familiar chart of the account balance here.

And you are absolutely correct, because looking at the original  Freqtrade plot and next the account balance plot. A similar situation  appears as the other higher timeframe strategies.

A few bad weeks at the beginning of the year, at around the lowest point of the bear market.

Which took your account balance underwater.

And then finally at the end of the year some lucky trades that skyrockets the balance again into positive territories again.

And with the Alligator strategy I can imagine this.

Since it is not fit for ranging markets, but is better suited for trending markets. Which started around Oktober / November.

And saying this made me think that I know whats going on with these performances until now.

The first three higher highs in the crypto happened so quick that the  strategies I discussed so far could not respond well to these pumps.

They all make use of lagging indicators, and thus did not have the  possibility to respond quickly to these rapid changing market  circumstances.

On the moment they finally started to trade, they already entered the game too late.

Resulting in trades that lost or had no or very limited profits.

Maybe somebody else that views this post can confirm this or has their own theory on what the possible causes can be.

And if you do, please do add it to the comments below so everybody can benefit from your insights.

Let me finish this strategies performance with a comparison plot with the other strategies on their performance indicators.

And then move on to the next strategies placed at the 15th and 14th position.

15.Kijun cross strong spot and 14.Kijun cross neutral futures

Now  I told you that the Ichimoku indicator was one of my favorite  indicators and here I again have two strategy versions based on this  indicator.

Now for the people that do not regularlyfollow my channel.

One reason why I have so many strategies based on this indicator is  that I made a whole series on how to build your own algorithmic trading  strategy. And in these videos I used example strategies that were based  on a book based on the Ichimoku Kinko Hyo trading indicator.

So now you understand a little bit better why there are so many Ichimoku variants at the moment.

Anyway, If we compare these two results then we see that the end  balance of these two variants do not lie far from each other. The win  rate is also almost the same. But the amount of trades has a big  difference here.

The futures version has almost twice as much trades in comparison  with the spot trading version. Now I think this is normal since trading  with futures can be with long and short trades, thus having two  directions to trade on.

The thing that dissapoints a little bit here is that trading two  directions does not necessarily mean that profits are also double.

If I open the charts of the account balance we can see better how  their performance is over time and if there is much difference there.

Unfortunately the account balance charts do not look too optimistic.  But again I get another piece of information on why these type of  trading strategies perform like they do. Again it's too bad that the  futures version did not make optimal use of bullish and bearish market  moments. But instead has a sharper decline after the bots active trading  in comparison to the strong signal oriented spot version.

If you watch closely you can see that the strong spot version started  to trade at around April 2023 and the Neutral futures version at march.  The main reason for these late trades is that with these strategies the  Kumo cloud has to form first to  complete these indicators. Only after  this data is available there can be trades made.

The other, slighter difference, is that the neutral version can trade  earlier because it can start to trade when the close price is still in  the Kumo cloud. Whereas in the strong version we have to wait until the  close price is above or below the cloud. Maybe this is the reason it has  a higher drawdown right after it started to trade

But this is a detail between these two strategy versions.

Again we can see that only at the start of the third quarter of 2023  these strategies came to life. When the possibilities of long swing  trades came available. And finally starting to catch some major pumps of  altcoins.

For example with the spot strong variant it catched the AVAX, SOLANA,  CARDANO and some others altcoin pumps. Making the losses of the  complete year good in a couple of trades.

Do I now have to conclude that these strategies are bad?

I think not.

But I can at least say that I start to understand the trading profile  of a swing trading strategy on higher timeframes better in relation  with the low winrates.

For these kinds of strategies it is almost a matter of keeping the  faith that at some point they will catch the trades that make all the  red numbers disappear from your account balance.

At exactly which moment this will happen is unsure and will certainly need more investigation.

But for me these results will not be a surprise anymore and I think I can understand them better now.

Unless somebody corrects me in the comments of course.

And please do by the way!

13. Flawless Victory strategy

The next strategy that enters the 13th spot is the Flawless victory strategy.

In the original strategy league this algo took the 24th spot on the  30 minute timeframe. But in 2023 it showed its strength and took the  13th spot in the test if you used it on the 1 day timeframe.

Not that it would defeat the Bitcoin baseline. But at least you would have a profit last year.

With 68 trades and a winrate of 70 percent over 52 percent of pairs used AND a maximum drawdown of only 10 percent or so.

Now this algorithm also had some sort of a warm up time and only started to trade around April or so.

It then would suffer the same crypto market conditions, where there  was a pullback after April's high. But after that is would have it's  best times and at least kept your account balance in profits.

So maybe I can start to be a little bit more optimistic from this  point on and will be very curious to see how the other 12 strategies  will perform.

12. Smart money strategy

On the 12th spot we find the optimized Smart money strategy.

This one was also placed as one of the lowest scoring strategies on  the original Strategy League. But now it scores pretty well on the same  15 minute timeframe.

However If you take a good look at the amount of trades, you see that  something is not right here. The best timeframe is the 15 minute  timeframe and it has only 14 trades, and with only 26 percent of the 50  pairs in the configuration.

This also results in a Calmar ratio of 337 and Sortino ratio of 134. Something is fishy here.

And indeed, if i open the profit curve's weekly performance on this  algo, then we can see that it actually does nothing the whole year.

And by nothing I do not mean it does not trade. But it opens trades on some pairs and keeps these open the whole time

As you can see in the paralellism plot.

In this case the first signals would come from XEC, QNT, ETC and some others.

At the end of the year these positions would be closed.

Effectively setting the parallessism to 0 again and also improving the account balance to 1600 dollar.

However, besides the fact that I do not really think this is trading  but more of an investing strategy, I think that, as long as you hold  these trades open you have the risk to lose your money.

Finally I suspect an error in this trading algorithm for some reason.  And I will not blame you for having the same thoughts when you see  this.

On the other hand. Maybe smart money does not trade that often and  only invests in real opportunities here. And this algo reflects that  opinion perfectly.

But don't let me fool myself. This is just very strange behaviour  from this algo and I do not think I would recommend to use this  algorithm unless I am absolutely sure what happened over last years  data.

Nonetheless, these are the results I get from the backtesting engine. And this is just a strange exception here.

I'll just have to keep this in mind for the future and will proceed with the next strategy on the list.

11. Kijun cross strong Futures

And  I warned you earlier. The Ichimoku indicator was tested rigorously and  had a lot of strategies entered in the original League.

But I promise, there is only one Ichimoky based strategy left. So bear with me...

This version does short and long trading and will only trade on the  strongest signals with the most confirmations. So actually the most  advanced version of the bunch.

However, It still suffers from the same problem as the other long term swing trading strategies on the list.

Initially it seems to have good numbers on factors like drawdown,  Calmar, Sortino, and Sharpe ratio. As well as the profit factor.

But looking at the way the algo performs over the year. After the  warmup the algo already suffers its drawdown as a consequence of the  market circumstances.

Luckily the drawdown only was slightly more than 12 percent.

And again around Oktober the algo experienced more positive weeks  than negative. Until the last week of the year, where the best trades  occured.

Because most strategies have these real good trades at the end of the  year I now also believe that this is no error in the tests but the  market circumstances favored these type of strategies around that time.

Not that it mattered much though, because until now all tested algo's  do still not come close to the initial baseline where a profit of 170%  was reached. Also the best Ichimoku variant does not come close.

So lets now start with the countdown of the 10 best strategies tested over the data of 2023.

10. Low Bollinger Bands strategy

At the 10th spot there is the lower Bollinger Bands strategy.

The end balance of this algo is not the highest of all the tests so  far. But overall the other numbers of the 5 minute timeframe do seem to  be balanced out. Except for the Sortino ratio which is exceptionally  high.

The profile the account balance has begins to make me somewhat  optimistic. There is only a drawdown of max. 5.5 percent and at the half  of 2023 this algo starts really perform well. With some good profitable  weeks.

I also start to notice that the account balance curve zigzags around  the 0 line more. Which also results in a higher score ofcourse.

Now a profit curve does not say anything without the context of other  performance indicators which also gives you insights in the risks you  experience.

But in relation to the strategies that are lower than position 12 you  now can clearly see that this performance line does not dive deep into  the negative profits and stays relatively close to the zero line or  higher.

Here The Low Bollinger Bands strategy scores primarily on it's high  Sortino and Sharpe ratio and it's low drawdown. Also the high profit  pairs ratio adds up to the total score.

09 BeastBotXBLR7

At number 9 we find the Beastbot strategy with an endbalance of 1300  dollar. Still way below what we could had have if we bought our bitcoin.

But I can see a new pattern emerging.

The winrate keeps reasonably high. Also in comparison with the amount of trades.

And the drawdown is almost minimal.

Pairs respond well and the ratios do not seem to be that out of range.

Taking a closer look at the profit curve, then we finally see a  strategy that gains profits right from the start and, although it has a  drawdown, it did not put your account balance under water.

Also whe wins and losses over the year give some satistaction since  you would experience more winning trades than losers. And that creates a  confidence boost that is not mentioned enough in trading psychology.

Overall this algo scores good over almost all performance indicators.

And hopefully the Beasbot strategy will be the first one that gives  us a glimpse of the performance of the strategies with their algorithms  that hold place 8 and higher.

So lets continue quickly with number 8 on the list.

08. Bollinger Bands Real Pullback TSL-BI

And  maybe I'm too optimistic here by saying that the pattern of positive  trading strategies gets confirmed here. But looking at these results, we  definetely enter the territory of algorithms that would not have  dissapoint us in 2023.

Again, I keep repeating myself, but the endresult is still not comparable with the Bitcoin investment.

But at least we might see the strategies that on the longer term,  let's say a next bull market cycle, that have the possibility to  outperform such a Bitcoin investment.

The interesting thing here is that this specific strategy, on the 30  minute timeframe, the amount of open trades is not that spectacularly  high. Instead the trades are short and profitable. And a coin like GALA  responds exceptionally well to this algorithm. As if it was made for it.

I just told you about the psychology of having more wins than loses. And here is another example.

Now you already know that this is inherent to the type of strategy.  Like swing trade strategies that have more losing than winning trades  and in this case a strategy that apparently has a better way of handling  ranging markets.

Like what happened in 2023.

You can see that clearly by looking at the curve of the cumulative  profit. The algorithm catches some good pumps at the start and end of  2023. And at the same time seems to hold well when markets are ranging  and is not or trades less in the ranging zone.

Preventing losing trades and a drawdown. You can see that pretty well  here because the wins and the losses lines stay relatively flat around  these periods.

THe plot that shows the average profits and loss over the year also  shows here that at the beginning and at the end of the year the highest  average profits were made.

Where the biggest drawdowns occured in the middle of the year. And  mind you, that by BIG i mean only a max of around 2 percent  here.

So overall this strategy would be a joy to see in action in 2023. And it definitely deserves the 8th spot.

Now, let's quickly see the 7th spot and see how that one performed over 2023.

07. Obelisk Ichimoku ZEMA

The Obelisk Ichimoku Zema currently holds the top spot of the Strategy league over the data pre 2023.

In other words: this algo performed the best over all the previous bull and bear market cycles of the Crypto I use in the list.

Therefore It surprises me a little bit that, if you would use this algorithm over 2023, it would not perform the best of all.

If you watched the video about this algorithm, you might already know that this one only works on the 5 minute timeframe.

And I could not test it on other timeframes.

Not that the current performance is bad.

But we still have to keep the Bitcoin investment in mind and this performance still lags behind.

But the profit curve is a beauty to see and it would still be a joy to see your algo attain profits at a regular base though.

Also, over 2023 it had an average drawdown of 1,5 percent. Which is  truly admirable to say the least. Better than the swing trade strategies  I showed earlier.

As a sidenote I also have to say that the moment you start a bot that  executes an algorithmic trading strategy is of the utmost importance.  Over the dataset of 2023 I had the luck that 2023 started with one of  the lowest points of the bear market and it ended almost at a years  high.

Would I do this test over 2022 then all strategies would perform  probably very poorly. Perhaps with the exception of the Futures version  of some of these strategies that can trade shorts as well.

So if you watch one of my video's or any other trading channels  strategy video. Then the moment you start to trade that specific  strategy manually or start a trading bot with that algorithm is also  important.

So if you see one of my testing videos and you start your bot with  that algo and it performs unexpectetly poor. Then watch the overall  chart and ask yourself if you started the bot at the right time.

So please keep this information in the back of your head if you want to start trading or want to start using a bot.

Now, let's get back on track and continue with the number 6 strategy on the list. Which is the BinHV45 strategy.

06. BinHV45

This 15 minute timeframe  algorithm has a 50 percent profit percentage. The individual ratios  numbers do make me slightly worried since they are very high rates. It  could be that this strategy has something to show us that needs further  investigation and explanation.

Since I also think that the amount of trades for this timeframe is pretty low too.

And looking at the way the cumulative profits grow there is something interesting going on between september and november.

It seems that this algorithm has a good indication on where the  market ranges and were it pumps. Because during the September / November  period the overall crypto market ranged. And from november it started  to move upward again.

And you can also see that on the other two flattish lines around April / May and again around June / July.

The trades chart also adds more detail to this story.

This algo has a mechanism that detects ranging markets and activates  again when the market moves again. You can see that more clear on this  plot because the amount of trades does not rise during these ranging  periods.

And that also results in a low drawdown that stays relatively stable  as well. You can see that with these flat lines here and there.

This also makes this trading algorithm one of the first that has a  very steady rising profit curve with hardly any big drawdowns.

Again, this is only over the market profile of 2023 and other years with other profiles can negate this conclusion.

But as for now I can honestly admit that the last couple of algos are  certainly worth my attention for the coming bull market. Even though  the profits are not that spectacular.

So with this number 6 spot discussed we are finally entering the top 5 trading strategies that could or should be used in 2023.

Again it is always 20/20 hindsight with these tests and a  hypothetical excersise nonetheless. But with this information we can  make more educated guesses for the coming future...

05. ClucMay72018

Now the Clucmay  algorithm does not make you the most money and the endscore is only  marginally better then the just discussed BinHV45 algo. But its  performance numbers are still impressive.

And the way the profit is build up does make me feel that I have a small preference to this strategy instead of the BinHV45.

The reason here is more subjective and has more of a personal psychological background.

I'd rather see a trading bot using an algorithm that is active and  increases my account balance ever so slightly but steadily. Instead of  having a bot doing nothing for a couple of months, making me doubt if  the algo still works and in the end trying something else.

While it was acutally the exact intention of the algo to do nothing  at that moment. Only it does not mention that and making me think it is  broken.

So here you can see that there are still trades happening, even  though some of them are losses. But the winning trade line is still  moving away from the losing trade line.

Making the endresult a winrate of 65 percent.

So you can see at the overall performance indicator scores that this  strategy is not always the best at some performance indicators.

But consistensy can also make you win championships, even though you never won a race or a match the whole season.

So this makes this trading strategy end up at the 5th spot, with a  very small difference in endscore between the numbers 5, 6 and 7.

04. The SMA Offset Protect V1 strategy

The trading strategy that I tested and landed on the 4th spot is the SMA Offset protection strategy.

It was the second best performing trading algo so far and with it's  hypothetical performance over 2023 it landed on the 4th spot.

And well, Which I think is not bad at all. The orginal best timeframe  of 30 minutes did not change. It still performs best at that same  timeframe.

And with a 37 percent winrate with a drawdown of less than 4 percent  this makes me again curious to see what the cumulative profit profile is  over the year.

Again these flat periods where there seem to be no trades at all for  weeks still makes me worry a bit. Would I keep faith in this algo if I  do not see any action for days or even weeks? Even on this low  timeframe?

Now that I see this profile I wouldn't worry about this anymore in  the future. But if I started a bot a year ago and then see this  happening I would not be so sure about it.

You can see it on the chart with the trades as well. Periods where there is no trade made.

Again these are the periods where there is a sideways market and I know this now. But again would I trust this in the past?

Anyway, The drawdown periods are sharp and last not that long. And  all occur on a positive account balance. So not too frightening.

And the comparison whith the other well performing trading algorithms does seem to have a similar pattern.

So let's just take a short look at the overall scores of the  performance indicators here and see if there are any remarkable things  here.

The only real thing that draws my eye is the win percentage of this trade. And if I see this well, it is the second best score.

In this overview i will not really count the score of the smart money  strategy here. Since it had so few trades and that can easily lead to a  very high percentage.

So if you put it into perspective maybe this Sma Offset strategy has  the highest winrate so far if you count only the active algorithms here.

The other graphs seem to show no spectacular things here at first sight.

So let me continue with the next two strategies on the Top25 best performing algo's over 2023.

And these are close siblings of each other.

Namely the Bollinger Bands Real Pullback TSL variants.

03.BB_RPB_TSL_RNG

Now I must admit  that I only made a video earlier of the better performing version of  this strategy. But I kept the other version as well in  my own league  list since I found it performing so well.

Also I do not always make Youtube video or post of all strategies I test since it would take up a lot of time.

And some of these algo's do have such bad performances that make the  invested time to make these video's and blogposts not worthwhile.

Nonetheless I keep the results because otherwise I might test them double later and that is also a waste of time.

Anyway, Both these Pullback strategies have almost similar and good  end results in the 2023 test. Whereas in the original strategy leagye  the're end results lay further apart.

Also this time they have better results on the 1 hour timeframe instead of the 5 minute timeframe in the league.

The V2 version scores better on profit percentage, winrate, average  and winstreak. But it lags behind in drawdown and ratios in comparison  with the other pullback version.

Nonetheless there are very minor differences and that is why they score that close together.

If I compare these strategies on their profit curve. Then we see that  on the weekly chart the second version has ever so slightly snappy  trading results with these small peaks. The first version is more smooth  and has also longer flat periods at first sight.

This can also be seen if I compare the profit loss charts with each  other. The second version has some higher performing weeks but can also  suffer from weeks that have higher losses as well. Making the average  profit line fluctuate more over time.

If I only use the performance indicator plot, then we can see better  at which indicators the second version of the Bollinger Real pullback  strategy perform better or worse than the first version.

Again the differences are not that big between these two and it's  only a matter of personal preference and your own deep dive  investigation to find out which version has youre preferred trading  profile.

However, there is one important point that I can conclude at this point.

Even though we landed on the second position of the best performing strategies I tested so far.

We still did not manage to make more money than the simple Bitcoin  buy and hold strategy that we used as our baseline for the year.

Trading warning

Does this mean these specific Top 25 examples make them bad strategies?

I personally think not because I tested these all out and they could potentially make money based on the results in the past.

However, in trading there is always the fact that past performance  does not guarantee future results. And all the tests I do are based on a  hypothesis and only tested on this historic data.

So You always should take into consideration that trading involves risk taking.

And you should be aware that you potentially can lose all the money on your account balance.

Therefore you should only use the money you are willing to lose and can afford to miss.

No matter if it is with manual trading, bot trading, investing and  other investment types like NFT trading, gambling and other stuff where  your money is at risk.

Even setting your money aside on a bank or a savings account involves  risks, although not that high. Because your money can devaluate because  of inflation.

It all has to do with your risk profile and the willingness to put your money at low or high risks.

And now without further ado, lets see what the best performing algorithmic trading strategy over 2023 would be.

In theory we had to only use this one trading algorithm to attain the highest profits over last year.

And that is...

01. The God Strategy

The Godstra algorithm.

It is all in the name.

The Godstra algorithm seems to had the best edge to provide us with the greatest wealth in 2023.

What also strikes me is that this is the only strategy that actually  outperformed the Bitcoin buy and hold investment over 2023 as well, with  a theoretical profit of 252 percent.

This wealth was attained on the 1 day timeframe. But even if you had  used the original best performing 4 hour timeframe, you had outperform  the Bitcoin baseline.

This trading strategy has a staggering 97 percent win rate, but also with a reasonably high drawdown.

However, to be straight to the point and realistic here, the 22 percent pairs rate catches my eye immediately.

This indicator tells me right from the start that only 9 or 10 pairs respond positive to this trading algorithm.

All the other values look to be quite normal in my view.

When I open the logfile of this timeframe I see that, in this case  the algorithm heavily depends on altcoins and their tendency to pump  heavily.

And if I show the profit curve.   We can see that this algorithm pumps heavilty at the beginning of 2023.  When the market rallied to its first high in that year after a dreadful  bear markets low.   As it seems now only some of the the altcoins we've seen an couple of  seconds ago triggered this algo to make a buy.   And only a few of these pumps made the initial profit end up at more  than 1000 dollars profit. So these few trades were better than a whole  years worth of trading of the second place scoring algorithm.

After that a period of consolidation began and during that phase it exerienced its drawdown peak of 27 percent.

Now mind you that was a peak that was way beyond the average drawdown of around 8 percent.

If I want to see the way the win percentage of over 88 percent was build up. Than I will just show the next plot.

Now, It almost seems that there are no bad trades here. But then  again this algo only traded 10 pairs and only trading those pairs was  enough to outperform the Bitcoin baseline.

Another thing I want to mention with this strategy is its active trades.

The following plot shows an overview of some of the other plots I already shown you. But in another form.

But this time I want to focus on the Paralellism plot.

Here you can see that the bot is active the whole time. Most of the  time holding a position or quickly exchangng one coin for the other.

Some of the pairs that also performed well are AMP and GALA. Together with some other well known altcoins.

And especially these altcoin pumps made the algo perform like we know now.

To round this off I only want to show you the overview of al  the performance indicators again.   And its funny to see that the best performing strategy, with its best  profit and win percentage, suffers the biggest drawdown peak of them  all. Together with the lowest pairs with profit percentage.  This all seems very contradictionary but also logical at the same time  if you have spend time enough in crypto.

Because my personal experience, and that of other crypto youtubers  that you might follow, is that the biggest gains is with altcoins and  not so much with bitcoin.

The king of crypto is worth your investment if you look at it from a  long term perspective with a reasonable amount of risk. And it should be  the base of your portfolio.

But the biggest gains in any crypto market comes from the altcoins.

However you will have to be carefull not to buy randomly. Because  there is a high possibility that you will end up being a bagholder of  worthless coins.

Cherrypicking or having a strategy on which pairs to trade is of the utmost importance here.

In my case I used the top 50 best selling pairs of 2022. But you can  also think of a diversifying strategy with, for example trading the top  10 coins based on market cap, on their level 1 puroose, NFT platform,  gaming plaform and so on.

The possibilities are limitless here and all depending on your risk appetite and familiarity with the crypto market.

I just wanted to demonstrate here that there are no absolutes in trading and there is nothing sure in the future.

With the knowledge gained from this excercise it could be a good  overall investment strategy to buy and hold A couple of investment  worthy coins like Bitcoin and Ethereum. And besides that use trading  strategies that aim for long term swing trading and shorter term  scalping. So to diversify your crypto investments and thus spreading  your risks.

Wether it be by doing manual trading or using a bot to do the trading for us.

But for now I think I have made my point. And I must prevent myself from rambling on about this topic and keep repeating myself.

You have just read an overview of a test of the top 25 best  trading strategy alorithms for crypto and how they would perform over  the year 2023.

What you see here are the results of this video's 2023 test. When I  will continue to test out future trading strategies and algorithms I  will use the old data again and thus compare over another timeperiod.  But I already explained the reasons for this earlier. So do not be  surprised to see the old league back again in the next video I make.

I hope that you found this entertaining,  informative and usefull enough and I hope you learned something from  this video. I certainly know I did.

Have a succesfull next bullrun and  I will see you in the next post.

Goodbye