Menu

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Can someone backtest my Expert plus Setfile for... 2 replies

Can Someone help me to Code my Simple Strategy? 0 replies

Visual backtest result different from non-visual backtest 0 replies

Can someone help me convert this simple program to an EA 5 replies

How can i backtest a strategy on meta trader 3 replies

Attachments: Simple BTC strategy - can someone backtest?
Exit Attachments

Simple BTC strategy - can someone backtest?

  • Post #1
  • Quote
  • First Post: Dec 20, 2024 10:28am Dec 20, 2024 10:28am
full moon buy , new moon sell , any good backtesters on here wanna see what they think???
Attached Image (click to enlarge)
Click to Enlarge Name: 6.JPG Size: 65 KB
ill try myself but my backtesting is terrible
I do a fair bit of backtesting for algorithmic futures trading and have boilerplate code that makes it pretty easy to set up experiments. Depending upon the complexity of your system, I might be able to knock something out without too much trouble. Feel free to reach out if you'd like a hand.
Quoting azhen
I do a fair bit of backtesting for algorithmic futures trading and have boilerplate code that makes it pretty easy to set up experiments. Depending upon the complexity of your system, I might be able to knock something out without too much trouble. Feel free to reach out if you'd like a hand.
@azhen yea that would be great, im totally unsure what risk/ reward ratio would work, it wants to be a simple as poss, its only 2 trades a month , on the daily chart, and i think more of a scalp situation than a swing

just eye balling the previous few years I'd guestimate 65-70% win ratio so a 1-1 risk reward would do it, but thats just speculation, dunno if you work with set pip SL or it could be tested using previous support resistance, i have no clue, but what i do think s there is something in it

does any of that help?
Quoting iamhewho
{quote} I'd guestimate 65-70% win ratio so a 1-1 risk reward would do it, but thats just speculation, dunno if you work with set pip SL or it could be tested using previous support resistance, i have no clue, but what i do think s there is something in it does any of that help?
Sure - I have a variety of prebuilt stop strategies and most often end up using ATR or swing-based ratcheting of stops. Granted, most of what I do is intraday, but no reason it can't be applied to a different timescale.

Do you have the system well defined, in terms of entry and exit signals based on specific criteria? Once the entry criteria are coded, it's pretty straightforward to backtest with a number of different stop strategies and parameters (with all the usual disclaimers about overfitting the data).

One other potential hurdle is that I only just signed up (normally, I'm just using this site for the public calendar info) and don't have the clout for PMs, so depending upon how sensitive you are about your criteria it may be a challenge to communicate those privately. Sorry, had not considered that when initially replying.
Quoting azhen
I do a fair bit of backtesting for algorithmic futures trading and have boilerplate code that makes it pretty easy to set up experiments. Depending upon the complexity of your system, I might be able to knock something out without too much trouble. Feel free to reach out if you'd like a hand.
i dunno how your backtesting is done, but a good starting point might be to see if holding trade for 2 days is better or worse than holding for the entire moon cycle (18 days?)
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 7.JPG Size: 17 KB
Click to Enlarge Name: 3.JPG Size: 18 KB
Click to Enlarge Name: febuary B.JPG Size: 19 KB
Click to Enlarge Name: 2.JPG Size: 21 KB
its kinds looking like full cycle is better, but i dunno , maybe SL at previous support/ resistance, i dunno
Attached Image (click to enlarge)
Click to Enlarge Name: error.JPG Size: 25 KB
Quoting iamhewho
{quote} i dunno how your backtesting is done, but a good starting point might be to see if holding trade for 2 days is better or worse than holding for the entire moon cycle (18 days?) {image} {image} {image} {image}
Ah, your approach is literally tied to moon cycles? Initially I thought that was just a reference to how you were marking up the chart. Have to be honest, the likelihood of there being any predictive relationship between celestial mechanics and price action is... on the low side. But, it's also pretty trivial to code up, so why not? I'll poke at it and post an update later. If there are more specifics that I'm missing, definitely let me know.
Quoting azhen
{quote} Sure - I have a variety of prebuilt stop strategies and most often end up using ATR or swing-based ratcheting of stops. Granted, most of what I do is intraday, but no reason it can't be applied to a different timescale. Do you have the system well defined, in terms of entry and exit signals based on specific criteria? Once the entry criteria are coded, it's pretty straightforward to backtest with a number of different stop strategies and parameters (with all the usual disclaimers about overfitting the data). One other potential hurdle is...
my own criteria is buy on a full moon and sell on a half moon , im not sensitive about this at all, if this works all are welcome to it, im using the moon phase indicator on trading veiw , but one can also look at a moon phase calender for entires
Quoting azhen
{quote} Ah, your approach is literally tied to moon cycles? Initially I thought that was just a reference to how you were marking up the chart. Have to be honest, the likelihood of there being any predictive relationship between celestial mechanics and price action is... on the low side. But, it's also pretty trivial to code up, so why not? I'll poke at it and post an update later. If there are more specifics that I'm missing, definitely let me know.

the likelihood of there being any predictive relationship between celestial mechanics and price action is... on the low side. i would totally agree but here we are lol, watched some guy predicting crypto market movement down to the day, based on this and some other stuff, no really my bag, but i really wanna know, and looking at it , it look like there might be something there
Knocked together a quick test looking back over the past 5 years, and here are the initial results. Unless otherwise noted, all of these results are long-only, as you initially laid out in your description, but the cumulative summary also includes how an inverse short position would have worked. I included that because the strategy, in theory, avoids retracements (otherwise, just buy and hold) and that can be tested with short positions.

The things that immediately jump out are:

  1. Win rate is about even
  2. Max time to recover on the long side was >2 years
  3. Sharpe & Sortino ratios are low
  4. Divergence between long and short (inverted criteria) performance
  5. Total efficiency (entries and exits) is pretty close to the zero line

Based on this, my hot take is that bitcoin has been generally going up over time. It's more or less what you'd expect to see with arbitrary trades in an up market. There are some places where it catches some nice moves, and I can see those being enticing, but it calls the top of a move pretty frequently, too. In any case, here's some info to add to your consideration of the approach. Enjoy!

Attached Image(s) (click to enlarge)
Click to Enlarge Name: Chart.png Size: 80 KB
Click to Enlarge Name: Summary.png Size: 41 KB
Click to Enlarge Name: Total efficiency.png Size: 23 KB
Click to Enlarge Name: Net Profit - Long.png Size: 26 KB
Click to Enlarge Name: Net Profit - Short.png Size: 25 KB
Quoting azhen
Knocked together a quick test looking back over the past 5 years, and here are the initial results. Unless otherwise noted, all of these results are long-only, as you initially laid out in your description, but the cumulative summary also includes how an inverse short position would have worked. I included that because the strategy, in theory, avoids retracements (otherwise, just buy and hold) and that can be tested with short positions. The things that immediately jump out are: Win rate is about even Max time to recover on the long side was >2...

very interesting thanks for this, just to be clear, this was shorts on a new moon (every time) and buys on a full moon (every time)?, i wonder what would happen with a 12k stop loss then

like you say we would expect it to go up over time, although we have been in a bear market since 2019 more than we have a bull market,

anyway if there was or there is something to this, it wouldnt mean that moon phases effect the price of bitcoin.

it would mean that the whales/ elites/powers that be manipulate financial markets based on things like the lunar cycle which would give credence to many peoples theories, but just based on lunar cycles would be over simplified

like you suggested it seems t cal the top and bottom quite a lot, if you wanted to tinker a little more eg, 12k sl i would be interested, but if you have had enough exploring the crackpot idea i would understand lol

many thanks
  • Post #13
  • Quote
  • Edited 8:50am Dec 21, 2024 7:50am | Edited 8:50am
Quoting azhen
Knocked together a quick test looking back over the past 5 years, and here are the initial results. Unless otherwise noted, all of these results are long-only, as you initially laid out in your description, but the cumulative summary also includes how an inverse short position would have worked. I included that because the strategy, in theory, avoids retracements (otherwise, just buy and hold) and that can be tested with short positions. The things that immediately jump out are: Win rate is about even Max time to recover on the long side was >2...

P.S i only mentioned 12k stop loss based on this chart and maybe 25% stop loss on other ? i mean once we implement a stopl loss this could be nice profits no?
Attached Image(s) (click to enlarge)
Click to Enlarge Name: mu logo 2.JPG Size: 46 KB
Click to Enlarge Name: january B.JPG Size: 38 KB
  • Post #14
  • Quote
  • Last Post: Jan 13, 2025 6:50am Jan 13, 2025 6:50am
ive learned that not all moons are the same , it depends if we are in solar dominance or lunar dominance , but today and tomorrow are full moon, after which im pretty confident BTC will start rising nicely, coincidence?

if BTC isnt significantly up by end of week , i shall have egg on my face
Attached Image (click to enlarge)
Click to Enlarge Name: IMG_2590.JPG Size: 55 KB
0 traders viewing now