Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Basket Knights - A Basket Trading Round Table 1,311 replies

BASKET TRADE Management: Using multiple trade management EA's on basket trades 9 replies

Rekon-X Auto Bots Expert Advisors for Forex Trading 103 replies

TMS basket: Dashboard for a basket of MA-Filtered RSI Signals 135 replies

Grid, Burst Grid, Grid Risk and Grid Profit 1 reply

Attachments: Grid/Basket Trading Expert Advisors
Exit Attachments

Grid/Basket Trading Expert Advisors

  • Poll

  • How do you feel about grid trading?
  •  
  •  
  •  
  •  
  •  
  •  
  • Post #1
  • Quote
  • First Post: Edited 8:24pm Sep 13, 2024 7:17pm | Edited 8:24pm
I've been into grid trading for a while, but it seems like most people are against it. I believe this has less to do with the trading style and more with how successfully it's utilized.

I started this thread to discuss and share grid trading expert advisors and ideas.

For the past few years, almost every expert advisor I've created has had an optional grid feature. It has worked really well in trending markets, especially when combined with hedging.

The expert advisor attached, the JB-Basket, is the "holy grail" of my findings so far. It includes several grid features:
- Grid: Places a new trade every X pips in drawdown.
- Grid Candle: Places a new trade at the close of each candle.
- Grid Candle Trend: Places a new trade with every bullish candle (for buys) or every bearish candle (for sells).

It also has a trend-switching feature, allowing it to trade and grid in the opposite direction. You can select from up to 7 indicators to determine the trend:
- EMA
- SMA
- RSI
- Guppy
- TMA-CG-2024
- TMA-RSI-Bands
- Half-Trend

Here's a video going over more details of the JB-Basket:
Inserted Video


Feel free to share your grid/basket expert advisors and ideas, and let me know how I can improve the JB-Basket. You can try it with the default settings on EURUSD on the 15-minute timeframe.
Attached File(s)
File Type: ex5 JB-Basket.ex5   111 KB | 188 downloads
File Type: ex4 JB-Basket.ex4   162 KB | 227 downloads
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
Hi JB,

thank you for sharing.
In my opinion one function would be very nice in this basket trading:

Getting an equity stop. That means f.e. Starting an account at 10k with equity stop 500:
If equity reaches 10500, close all orders and next equity stop set to 11000 and so on.

this would be essential in may eyes.

BR
James
Quoting jblanked
I've been into grid trading for a while, but it seems like most people are against it. I believe this has less to do with the trading style and more with how successfully it's utilized.
Most grids are run without a hard limit on maximum drawdown and just keep going deeper and deeper until price reverses (or the account gets a margin call). The balance curve of a grid looks fantastic, which gives a false sense of success. What I mean by that, is that backtesting a grid is hard. It is not unusual for a grid to run successfully for a few years with a nice straight rising balance curve. The equity curve however, will have sharp "spikes" pointing down which are the drawdowns. Eventually, after some time there will be a huge spike that is too large and will cause a margin call.

One could tweak the parameters of the grid to avoid the margin call during the backtest interval and end up with a beautiful straight line rising balance line, but it is very hard to measure or estimate what the probability is that the drawdown will be too big on future data. You could try to backtest with thousands of random synthetic charts that were generated with the same characteristics as the actual symbol (same volatility, same trends, etc.) and see how often you hit max. drawdown. But synthetic data might not reflect the actual crazy stuff we encounter every now and then in the real markets.

Long story short: in my opinion, the only way to "safely" run a grid is by limiting the maximum drawdown. Stop digging in deeper and deeper after a given point and take your loss. Your balance curve will no longer look like a nice straight line, but more like an ugly sawtooth. But hopefully a sawtooth that slowly rises over time...
Quoting yoriz
{quote} Most grids are run without a hard limit on maximum drawdown and just keep going deeper and deeper until price reverses (or the account gets a margin call). The balance curve of a grid looks fantastic, which gives a false sense of success. What I mean by that, is that backtesting a grid is hard. It is not unusual for a grid to run successfully for a few years with a nice straight rising balance curve. The equity curve however, will have sharp "spikes" pointing down which are the drawdowns. Eventually, after some time there will be a huge...
I think what you described is mostly true if the EA keeps trading in the same direction. In almost all of those cases, you'll need to hedge or close trades in huge drawdown.

The trick I found was to allow it to switch trends and grid in the other direction as the trend changes.

Here's an example from Darwati: https://www.forexfactory.com/thread/...1#post14999751
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
Quoting jblanked
The trick I found was to allow it to switch trends and grid in the other direction as the trend changes.
isn’t the premise of a grid that you can’t predict the random markets and therefore just keep scaling in/averaging down until price moves in your favor?

Or put differently: if you can detect when the trend changes, why not trade the trend in a conventional manner?
Quoting yoriz
{quote} Most grids are run without a hard limit on maximum drawdown and just keep going deeper and deeper until price reverses (or the account gets a margin call). The balance curve of a grid looks fantastic, which gives a false sense of success. What I mean by that, is that backtesting a grid is hard. It is not unusual for a grid to run successfully for a few years with a nice straight rising balance curve. The equity curve however, will have sharp "spikes" pointing down which are the drawdowns. Eventually, after some time there will be a huge...
Hey yoriz,
Assuming what you say is true, any EA would have the same problems, not just grids, right? After all, I am only aware of one large investment fund that uses a completely automated EA to date. I believe the main issue is not whether the EA will fail, but how much you will have made before that happens. After all, if you can double your account 2 times, and lose your entire deposit 1, you are still in profit (assuming your deposit remains the same). Perhaps too much emphasis is placed on the need to create the perfect EA.
Quoting jblanked
{quote} I think what you described is mostly true if the EA keeps trading in the same direction. In almost all of those cases, you'll need to hedge or close trades in huge drawdown. The trick I found was to allow it to switch trends and grid in the other direction as the trend changes. Here's an example from Darwati: https://www.forexfactory.com/thread/...1#post14999751
Thanks for sharing the EA jblanked. I am currently testing a hedged grid EA, with completely random trades, that I created almost by mistake. This EA made $700 in 10 actual days of trading with a minimum lot, on a single cross. The bad part? I am sure it will explode, because it trades averaging down with incremental lots. Is it useless? Not for me, because I am learning how to limit the DD and because these are completely random entries. Eager to figure out how you keep DD under control
Quoting jblanked
I've been into grid trading for a while, but it seems like most people are against it. I believe this has less to do with the trading style and more with how successfully it's utilized. I started this thread to discuss and share grid trading expert advisors and ideas. For the past few years, almost every expert advisor I've created has had an optional grid feature. It has worked really well in trending markets, especially when combined with hedging. The expert advisor attached, the JB-Basket, is the "holy grail" of my findings so far. It includes...
To be honest, I don't know the term system that I use
but if you want to develop a system like mine,
Attached Image (click to enlarge)
Click to Enlarge Name: XAUUSDM5.png Size: 49 KB


you must have the concept of several close orders according to conditions
for example:
* close the order because the profit is > a certain percentage
* close order because there are too many orders
* close order because the price exceeds the upper/lower band
*close order due to time
*etc

It's very easy to open an order, but difficult to determine when to close an order
Honestly, it just takes courage to try it, follow the process and let time decide

JBlanked, good luck
Quoting yoriz
{quote} isn’t the premise of a grid that you can’t predict the random markets and therefore just keep scaling in/averaging down until price moves in your favor? Or put differently: if you can detect when the trend changes, why not trade the trend in a conventional manner?
Yeah, I think that's how most people intend to use grid trading.

As for your question, "Why not trade the trend in a conventional manner?"—grid trading allows you to potentially gain more profit if the market moves against you and then returns in your favor. Since we can only predict future trends, it increases your odds of success.

The key advantage is having it switch trends. I’ve seen that it significantly lowers drawdown.
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
Quoting Fiox89
Hey yoriz, Assuming what you say is true, any EA would have the same problems, not just grids, right?
No, "conventional" EAs that have only 1 position open at a time can be properly backtested. The equity curve will not be a nice straight line, but is representative for the behavior of the strategy. Assuming it is not curve fitted, you can estimate what drawdowns and profits to expect.

With grids however, you will with get a straight line equity curve.... until you don't. Everything looks great and BAM! suddenly your account is gone. This is very hard to backtest, because it is trivial to tune the parameters of a grid to avoid the big drawdowns but you are just kidding yourself.
Quoting Fiox89
{quote} Thanks for sharing the EA jblanked. I am currently testing a hedged grid EA, with completely random trades, that I created almost by mistake. This EA made $700 in 10 actual days of trading with a minimum lot, on a single cross. The bad part? I am sure it will explode, because it trades averaging down with incremental lots. Is it useless? Not for me, because I am learning how to limit the DD and because these are completely random entries. Eager to figure out how you keep DD under control
Thanks for sharing and joining the conversation.

In my experience, incremental lots work well if you have a hedge. The trick in that scenario is determining when to hedge, which is honestly the hardest part.

I’ve also experimented with some "outscaling" techniques, where you close the initial orders when the profit from a portion of the later order(s) equals the profit of the entire initial order(s).
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
Quoting yoriz
{quote} No, "conventional" EAs that have only 1 position open at a time can be properly backtested. The equity curve will not be a nice straight line, but is representative for the behavior of the strategy. Assuming it is not curve fitted, you can estimate what drawdowns and profits to expect. With grids however, you will with get a straight line equity curve.... until you don't. Everything looks great and BAM! suddenly your account is gone. This is very hard to backtest, because it is trivial to tune the parameters of a grid to avoid the big drawdowns...
I think you're speaking from your own experience, but that hasn’t been the case in mine.
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
brother this is,,, trend side trade,,,, not grid type trade one time one trade,,,,

i told one shot ea,,,,,, but with trend only
Attached Image (click to enlarge)
Click to Enlarge Name: Sell call.jpg Size: 166 KB
brother i think you not understand ,,,,
Quoting arun21385
brother i think you not understand ,,,,
This thread is about grid trading brother. For coding requests, use this thread: https://www.forexfactory.com/thread/post/15000224
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
Quoting yoriz
{quote} No, "conventional" EAs that have only 1 position open at a time can be properly backtested. The equity curve will not be a nice straight line, but is representative for the behavior of the strategy. Assuming it is not curve fitted, you can estimate what drawdowns and profits to expect. With grids however, you will with get a straight line equity curve.... until you don't. Everything looks great and BAM! suddenly your account is gone. This is very hard to backtest, because it is trivial to tune the parameters of a grid to avoid the big drawdowns...
Quoting jblanked
{quote} I think you're speaking from your own experience, but that hasn’t been the case in mine.
In general, I have always had little faith in backtest reliability, but what about comparing backtest results with forward test results of the same period? Would there be any benefit in knowing that the results are very similar? I did this test, and found that the results are very similar, although the EA in question is very simple and perhaps this makes the backtest more reliable.
Quoting Fiox89
{quote} {quote} In general, I have always had little faith in backtest reliability, but what about comparing backtest results with forward test results of the same period? Would there be any benefit in knowing that the results are very similar? I did this test, and found that the results are very similar, although the EA in question is very simple and perhaps this makes the backtest more reliable.
I complete agree; forward testing (live trading) and then comparing backtest results ensures the advisor is trading as intended.
Software Developer and Music Producer
News Fundamentals All Time Return: 1,159.1%
  • Post #20
  • Quote
  • Edited 10:00am Sep 17, 2024 9:48am | Edited 10:00am
Quoting Fiox89
what about comparing backtest results with forward test results of the same period? Would there be any benefit in knowing that the results are very similar?
Yes, that is important. You need to know whether your backtest software is properly simulating trading costs, spread, slippage, latency, etc. Platforms like MT5 are good enough for most retail strategies. Only for quick HFT scalping the MT5 backtester is not ok.

However, even if the tool perfectly simulates the market, it is still important how you use the backtester tool. It is all too easy to overfit a strategy. You can optimize input parameters to create beautiful equity curves in a backtester that avoided all the problematic news announcements, price spikes, etc. but will perform poor on the future live market. So don't use the backtester for (over-)fitting input parameters, but only to get a feel for the behavior and typical drawdowns and profits you can expect.
0 traders viewing now