Casino EA Trader

Ichimoku for the win today! Exposure is way to high for the account size, yeah and it's only demo So moving forward going to let him keep running with no EL, and check in on what the filters would have looked like. I imagine Heat-map and or % Change will work well with this one as it trends, however maybe just limiting the number of trades. Wellp any given week day, stay safe and enjoy.

Hi Alula, Yes controlling risk on a micro accounts is really difficult even if the minimum lot size is small like 0,01...

This is one of the single biggest reasons why so many people fail,.. they start underfunded, i.e. starting account size is too small..

Yes, I know that many people do not have the luxury to start with 10k or more, I completely understand this,.. But if you want to ensure you have a reasonable shot at making a success you need to start with at least 10K and 20K will be even better. Believe me, this is not easy for me as well, but i know from experience that this plays a big role.

regards...

P.s. Be safe wherever you are guys..

H

I absolutely agree, this happens to be a demo account I have run down over time. Still fun to crash them from time to time

I am interested in your risk management experience from your hedge fund days. In my current live trading accounts I do not risk more than 2% per currency and scale out 50% at 1 ADR then wait for an exit indicator to claim the rest with an emergency SL at 1.5 ADR. However if trading for someone else and you are not aloud to have a large draw-down (-4% from your highest watermark) this technique can be problematic unless you lower your % per trade. I am currently trying to incorporate and measure how I will control this in the dashboard, my current solution is to limit trades and/or % per trade.

I recall you mentioning Gross Exposure over 500% is not advised, do you have any other money management recommendations that can be used along with the dashboard? I firmly believe that proper money management over time is the key to success, just trying to figure out how to match safety up with the dashboard there are so many options.

Hi alula

Man, oh man... Risk Management (coupled with Money Management) is a field of study on its own,.. and I can talk about this for hours.

So, yes, you are absolutely right,.. as a rough thumb suck you should never risk (risk to lose) more than 3% (but typical 2%) of your portfolio on any given trade. This is the simplest and easiest way to do risk management if you are not into quantitative analytics and dont want to break your brain, but if you are willing to go down the rabbit hole there is a fantastic method called the Kelly Criterion (https://en.wikipedia.org/wiki/Kelly_criterion) which can beautifully solve this question for you.

The Kelly Criterion, tells you what the optimum risk fraction of your equity should be, per trade, in order to have the largest compound growth rate while ensuring you dont blow yourself up.. What you do is you take the closed out trade returns for at least 30 trades (but preferably 100) for a specific strategy and apply the data to the Kelly Criterion in order to derive the optimum risk fraction per trade for that strategy. This is what we used to do at the Hedge Fund and so we continuously optimized our risk fractions based on hundreds of closed trades in the recent past. Its quite a thing to manage, but we did it and it works incredibly well.

The Kelly Criterion was developed at Bell Labs in the 1950's (if i correctly remember) and was developed there to determine how much data to send over a copper wire before the data quality starts degrading. Yes, data starts to degrade at high speeds and so you need to control this. The criterion helped them to solve that problem.

Ed Thorpe (https://en.wikipedia.org/wiki/Edward_O._Thorp) was the first guy to apply this to black jack with immense success.. and later the wizards in the fund management industry picked it up and ran with it. The Kelly Criterion is great if you know what you are doing and have a very stable trading system with a lot of data.

If you want to know more about money management read the books published by Ralphe Vince. He has written two incredible books that are not for the fainthearted, but they cover Money Management in a spectacular way)

Back to 2% risk,....

the reason that it is important to stick to 2% as a basic rule is that if you go and study the law of large numbers and statistical probabilities of "runs theory" you will quickly understand that it is quite possible to have a run (a series of consecutive trades) of 12 consecutive losing trades that are stopped out. Believe me,.. in the 10 years in the Hedge fund, where i worked, I have seen this with my own eyes happen twice a few years apart. It can happen and it will happen !! at some stage in your life this will happen and if you are not controlling your risk it can absolutely blow you up. Go and calculate for example 12 consecutive, 5% compounded losses and see what your equity looks like afterwards.. its frightening how much you can lose of your account equity in a few stopped out trades and due to the asymmetry of positive and negative returns you need to make 100% return to cover a 50% hole in your account. its very painful.

If we did not apply very strict risk management rules, we would have blown up easily. During those two events i saw several other very large hedge funds blow up because they did not control their risk, but due to incredibly strict rules that our brilliant CIO put in place, we were ok and we continued to perform incredibly well for many many years.. Risk management is your friend and ensures longevity.!!!!!!!!!!!!


95% One Day VAR:

The other KPI that we used to very carefully heed was what is called the: "95% One day Value at Risk" and ""95% Ten day Value at Risk"" measurements. These are standard risk measures in the Hedge Fund Industry. Basically, take your portfolio of trades that are on the book and you calculate the daily % return for the portfolio for the last 100 days (as if you had these trades open for the last 100 days) and rank the returns form positive to negative (largest to smallest percentage values) . Then you calculate the 5th percentile of these 100 values. That final value tells you what size of "large negative event" in your portfolio you should prepare yourself for. In other words you can expect that loss (negative event) or larger to occur 1 in every 20 days with the current portfolio on book. So depending on what TF you are trading this method shows you what "bad case scenario" to prepare yourself for and I must say its pretty accurate. It pretty much hits that 1 in 20 event once in every 20 period time frame and that value is what you need to control. Note that i did not use the word "worst case scenario" but i used "bad case scenario" because the worst case can always be larger that what you have ever seen. but if you apply the "95% One day Value at Risk" method you will be safe and you will not experience a "Worst case Scenario". it will keep you away from this.

These values ( "95% One day Value at Risk" and ""95% Ten day Value at Risk"values) we watched with a hawk's eyes and we never allowed our portfolio to have a 95% one day VAR < -5% and also we did the same calculation for 10 day returns and never allowed the 95% Ten day VAR < -15%. If your portfolio shows "95% One day Value at Risk" values lower than the allowed limit , then you reduce your "Gross Exposure" in those assets in the portfolio that cause you to be over the limit. This is the golden standard in the Fund management industry. Below is a snapshot of a basic example.

Maybe if Seller9 is willing, I can guide him how to do this and he can build it for us and reflect it in the Currency Trader DB underneath the Gross Exposure reading so you guys can see what "large loss event" to expect 1 in 20 periods for the trades that are on book. Its very very enlightening and can help you a lot to fine tune your risk.. Seller9 can set it up for us in such a way that it is calculated based on whatever Intra day TF we choose. (e.g. 1H TF) and therefore the 95% one period VAR will tell each of us what "large loss event" to expect in the next 20 candles for whatever TF we selected. Seller9 can also then add some extra functionality where we set a cut off (e.g. a 95% one period VAR cutoff = -2%) and once that limit is reached in the DB the DB cannot open any more trades and waits for trades to close before it continues to open more.. This way you will have a great deal of control.

Portfolio optimization:

Im not even going to talk about this, because here it gets crazy complicated and there are hundreds of methods:
- Efficient frontiers,
- Sharpe optimal portfolios,
- Omega optimized portfolios,
- Downside deviation optimized portfolios,
- Markowitz mean variance portfolios.... Blah blah blah...


Gross Exposure:

The Gross Exposure is the second most important KPI and we were typically not allowed to go over 500%,
This is stock standards in the Hedge fund industry. We managed funds for 7 large institutional clients (large banks) and some of them did allow us to go over 500% but it was in rare cases and only with special disclosures etc..

The reason they allowed us to go to 500% and over was the fact that our fund's return stream was totally orthogonal to the market and even compared to other funds. In other words our fund generated returns that were completely uncorrelated to the broad market and other funds, so our fund was a great form of achieving diversification and helping them to reducing risk in their own fund.

What the large institutional clients used to do was run a "fund of funds" i.e. they invest in a whole range of different fund managers (like us) and would "blend" the different fund managers return streams. They effectively used our hedge fund return stream to reduce their "95% One day VAR) for their own portfolio (their "fund of funds"). So these institutional investors would typically have 60% in Equity funds, 20% In Bond funds and 20% in Hedge Funds and the Hedge fund component was to reduce their overall risk.

In general when a hedge fund start going over the 300% Gross exposure level, then institutional clients start lifting their eyebrows and you gotta do a lot of explaining. But because our return stream was so orthogonal they allowed us to have higher than normal Gross Exposure levels (run at 500%). They used us as a risk management tool in their Fund (fund of funds) and we got paid very very handsomely for this..

so there you have it.. All in a nutshell..

ok guys enough rambling again.. The wife is looking at me and i need to go and sleep....

regards

H

Thanks! I really love this kind of stuff It's a good thing I'm stuck at home until further notice so I can spend some time unpacking this golden nugget

Unreal! Too many winning trades, must have been a lucky night.

My results:

Hi Friends, hope everybody is safe.

I am now more focused on the Ichi+HM240+Rsi60 and I have a question.
I’m not using the TS, its just to show the trades on this example.

Attached Image (click to enlarge)

As you can see the profit [green] was just taken back [red] by the mother ship lol.

My CT is on the D1 TF.
this means, the new trade happened 24 hours after the original trade was triggered.

Any thoughts how we can avoid this scenario?

Hi Lanapips.

Technically there is nothing wrong here, both of these trades are valid trades.

I think you are trying to stay in trades too long, You definitely need a TP to lock in profits. That is the success of the HM240 and RSI60 i have been testing for 6 months now. Its the TP that enables you to lock in consistent profits.

and you definitely need an SL to keep you out of trouble.

I typically use TP = 0,4 and SL = 0,3.

If a FX trade moves against you by >= 0,3 ADR (i.e. SL = 0,3) then something is really wrong and the reason you got into the trade the first time is not valid anymore. The SL enables you to cut bad trades quickly before they burn you..

regards

H

Personally I would use tighter TP/SL settings to prevent this sort of loss. That reversal though was hardly unexpected, with a two days of strong downtrend some sort of pullback/correction would be expected. The thing with Ichimoku is it is normally used to trade longer term trends (4hr+ charts), the current market is very unstable and trends are virtually non-existent or collapse quickly, so don't expect too much in the current market.

yes sir, i actually have TP and SL employed on this system.

last week, i had good success on my TP/SL of 0.4/0.5.
i tried to duplicate it this week and success rate is not as good so far.

i like your explanation about why 0.3 ADR is a better SL and i will make this change right away.

thanks again for your inputs.

hi Uncle,

i will be employing a tighter TP/SL per H's explanation above.

that's a very good observation about the Ichimoku, in my log notes now.

i agree with your comment that the current market as very unstable, this is also the reason why spreads are very high.

thanks for your response.

Hi, in the current dynamic market conditions maybe instead of using a trend based indicator like Ichimoku I expect we should get better results using a swing based indicator like "Swinger" with "CCI" confirms (I don't recommend the Stochastic filter as it appears to be faulty). I just started trialling such a setup here is today's Asian market result:

Hi JustUncleL

A very very very valid argument. !!!

I was actually thinking about an indicator these last few days that would help us identify which mode (trending or sideways) the general FX universe of 28 FX pairs are in. This indicator would measure the extent of overlap between 1H candles. I (and you guys im sure) have observed for several weeks now that there is a shedload of sideways action (intra day on 1H candles) due to uncertainty and excess vols. i.e. there is little trend and lots of sideways action. see snapshots as examples

If "extent of overlap" is big then swing and oscillator based systems will do much better. If "Extent of Overlap" is low then trend following system will do better.

Soo, we all need to learn to be able to detect these two forms of market conditions and adapt our trading strategies accordingly.

Maybe Seller9 can somehow build us a "extent of overlap" indicator that can help us monitor these two main market states which will help everyone to make this decision.

regards

H

This would be the Holy Grail.

The problem is all indicators use past data, and we can not trade the past and we can not see the future, regardless of what some people selling software say.

Once a range is found from past data it is nearing the end, it then becomes a BO trade.

Old charts are wonderful for those selling trading plans as they can find all kinds of examples to show how wonderful their software would work, But try to find those same entry points at the right edge of the chart is a whole different story.

If my crystal ball hadn't broken I would be able to do this Holy Grail Indicator, but for now, I will just keep wandering in the wilderness. LOL

Everyone Have a Nice Day and stay away from the virus.

The Stochastic Filter works for me but it is very selective.

Here it is with the default settings accept for the levels are 30 and 70.

Mr. Ichimoku owning it again today!

noted. thanks Uncle.

OK I will have another look at it, but when I tested it last week it would either show all 28 pairs NT and then from time to time all 28 pairs B or S. All the other filters I have tried seem to work fine, I may have a faulty copy of S9_Currency_Trader?

With the current market being so dynamic I tried a slightly different approach yesterday, looking more at swing type trading.





Dual RSI is struggling again this week


RSI(60,2,5,80,20)+HM240 is completely failing