my journal 2

[Yamato]

Ok, I am done with all my head-scratching and so on.

The thing with the investors is on hold for now, for an entire month. In the meanwhile, in case we restart, I have asked them to include a long time friend of mine for the future, but I don't even know if this friend is interested any more, given the sharp drop in the equity line that we saw lately and which I will post tomorrow, updated to the last trade.

I don't even know if my father will be interested any more. It's like the banks, who give you the umbrella when it doesn't rain and take it away when it starts raining, or the swiss banks with the north african dictators, freezing their assets as soon as the dictators actually needed them.

All these people were willing to give me money when I didn't need it and couldn't take it, and now they may not be willing to give it to me any more.

The ideal partner would be this guy running the server, the long time friend. Yeah, because I don't like to show my systems to any more people than I already have, and this guy has them on his server, so I would not be showing him any secrets that he doesnt already know. Besides, he's a programmer and a networking expert.

Now, the only thing is that I have a commitment, so I have to make it work with all of them together, which would be best for me and for everyone. But what might happen is that no one will want to work with me anymore. Not even my father.

In that case, I'd take out a loan, and do things by myself again. Or, you know, make a little money on the side, robbing banks and similar. Then I'd invest it, from the prison. They might let me trade in prison.

But seriously. I am not desperate. I've been worn out so much during this year that now I am almost feeling relieved.

In case this thing stops for good, with the present investors, the first thing I am going to do is get rid of, by putting them in an "unused" folder... get rid of all those useless files they made me create, which have pretty much paralyzed my creativity. Of course: if every time you want to create a system, you need to update 40 sheets, then it's all over.

Not just that. The maintenance of this stuff has become an enormous burden for me. Each week I have to update 4 workbook, with 20 sheets or more. I got ****ing octopus coming out of my ****ing ears. I am Tony Montana, a political prisoner from Cuba, and I want my ****ing human rights.

If there's something that I hate is creating kilos and kilos of infrastructure, that gives you a fake impression of reliability and precision and instead just paralyzes you.

Why are people using sharpe ratio, sortino ratio and all that crap? With such unreliable subjects you should not give yourself a false impression of reliability by using all these metrics, that just paralyze you.

You should just use profit as a metric. How much money did such system make in the forward-testing? **** everything else.

**** margin.

**** Sortino.

**** Gaspar Gomez and **** the ****ing Diaz Brothers. What did they ever do for us?

After 3 weeks swimming, of healthy life, I am starting to feel all those pains in my neck, due to leaning forward with my head, towards the screen, which I can hardly see (we now have smaller screens at work, and I've had a laptop for over a year). I cannot last another year at work plus at home with all this time spent at the computer.

There is no way I am starting a new partnership with just one party this time. And I am not starting a partnership for just peanuts. I am starting a partnership in order to quit my job very soon. Unless it is the old-time long-time friend, or him together with the present investors, this time I will start 10 or more different partnerships at the same time, and make a ton of money from all of them, but I will not disclose my systems to any one of them. Then I will get my own capital and then I will invest it, right away. Not giving my systems away, ever again.

 

[bbmac]

Much as the market treated you like a Ginger headed step-child on that final day of trading (to below the uncle point,)...re any new investors; I think you may soon be finding out whether the line '..but I know what I did wrong...' holds any weight !

G/L

 

[Yamato]

I am going to have to start ignoring your negative posts. But they're intelligent, respectful and well-thought out so I am not placing you on my ignore list. Keep posting negative stuff on me every other post, as you've been doing, and I will start ignoring you, but I will not place you on the ignore list because you're still being respectful and providing intelligent content. Your dosage of negativity towards me and my work has always been a bit too high for me to take. But not so high as to place you on my ignore list and prevent you from posting.

"But I know what I did wrong" is not my typical line, which is the way you make it sound. It's not a line that I've been repeating at all. Once again, I am feeling like saying that you're manipulating reality to misrepresent me and downplay my work. I won't say it, because you say you're in good faith. So I will say instead that you're misunderstanding, constantly misunderstanding me, or even better: you're imagining negative things about me which I am not saying. The list is long and it starts with you saying repeatedly "some of which work in forward-testing" about my systems, instead of what I had said, which is: "virtually all of them are profitable in forward-testing". That is a big difference, just like this of depicting me like someone who keeps on screwing up and saying "but I know what i did wrong", and the whole sum of your posts doesn't sound like it can be written in good faith, but like a constant effort to subtly discredit me. It sounds like you're subtly spitting on my work. In a respectful way, of course, but I still feel offended.

 

[bbmac]

1. Re your post #3061 above: '...Not giving my systems away, ever again....' ....with your last investor (s) did you have to reveal your systems ?

2. Re: That big gain last Friday/ gig loss on the final day - are you able to say why the gain and subsequent loss was so large compared to say that previous -10k that occurred a week last Monday ? Was there a system that lost the majority of it ?

Please answer in your own time.

G/L

 

[bbmac]

Don't be offended, I am simply asking how effectively you feel you may be able to 'sell' to a new investor (s) that sees the Equity curve (which at first glance may put them off) - what went wrong and what steps you have taken to fix it.

For the record and the avoidance of any further doubt (believe me/don't believe me - up to you) I am impressed by what you have achieved to date, notwithstanding the recent events. I clearly differ from you in what I think are the Weaknesses/Threats to your project or at the very least the relative importance of those weaknesses/threats, and I have discussed them here with you, sometimes persistently because I do not wish to see you fail. I believe the underlying cause of the recent EC fall off to below the uncle point are contained in the 5 main points I (and to an extent one or two others) have been discussing with you here for some time.

If you are unable to enter into positive intelligent dialogue without feeling hurt by the challenges that presents then this may impede your chances of working with a your new 'proposed' partner ? which by the way, I personally think is a good idea !

G/L

I am going to have to start ignoring your negative posts. But they're intelligent, respectful and well-thought out so I am not placing you on my ignore list. Keep posting negative stuff every other post, as you've been doing, and I will start ignoring you, but will not place you on the ignore list because you're still being respectful and providing intelligent content. Your dosage of negativity towards me and my work has always been a bit too high for me to take. But not so high as to place you on my ignore list and prevent you from posting.

"But I know what I did wrong" is not my typical line, which is the way you make it sound. It's not a line that I've been repeating at all. Once again, I am feeling like saying that you're manipulating reality to misrepresent me and downplay my work. I won't say it, because you say you're in good faith. So I will say instead that you're misunderstanding, constantly misunderstanding me, or even better: you're imagining negative things about me which I am not saying. The list is long and it starts with you saying repeatedly "some of which work" instead of what I had said "virtually all of them are profitable". That is a big difference and it doesn't sound like this can be done in good faith. It sounds like you're spitting on my work. In a respectful way, of course, but I still feel offended.

 

[Yamato]

I don't have to fix anything. The systems' selection has a big drawdown. The large majority of systems being traded is profitable. The mistake was trading them without being ready to take the drawdown. There is nothing to fix. I'd trade my systems myself if I had the money. And I am not selling anything. I put all my efforts into substance rather than appearance, and as a consequence, I do not engage in the activity of "selling".

 

[Yamato]

1. Re your post #3061 above: '...Not giving my systems away, ever again....' ....with your last investor (s) did you have to reveal your systems ?

2. Re: That big gain last Friday/ gig loss on the final day - are you able to say why the gain and subsequent loss was so large compared to say that previous -10k that occurred a week last Monday ? Was there a system that lost the majority of it ?

Please answer in your own time.

G/L

These are short questions, that I can answer immediately.

Yes, I had to reveal my systems from the first day. I won't do it again.

Bad balance of systems. Silver, gold and oil have too much weight in the present combination. So when they trade, this stuff happens. Silver trades cause you a plus 10k and minus 10k per trade. Gold and Oil cause plus 5k and minus 5k per trade. The other choice was not enabling silver, which seemed like a pity, because the systems were very good, and still are. But we weren't ready for it.

 

[counter_violent]

Hello Travis, I won't bore you with commiserations ref the systems trading hiccup...which I am sure it is, just a hiccup.

Some random thoughts. Maybe some use....I don't know !

When increasing trading capital x 4 dont increase trade size x 4. The idea behind increasing capital should be to make the job of making money easier ( less pressure on the pot ) and not to make more money with what amounts to the same degree of risk.

LTCM was sunk by a previously unforseen combination of events, so no amount of testing will ever show up a new combination which cannot possibly be known in advance and that has never been seen before. (Black Swan Event / Multiples Thereof) Therefore a more conservative trading approach will usually win out in the long game. If nothing else, it will keep you in the game for a longer period and provide you with more in the way of data and stats to analyse and digest.

When Whole Market / Fundamental Re-alignment moves are occurring their should either be a circuit breaker in place to stop all trading or an equalization mechanism that kicks in which in effect neutralizes the drawdown beyond an acceptable level. Some sort of temporary hedging of the worst offending systems maybe. The worst that could happen in this case would be some extra cost in commissions / spreads etc, but at least capital would be preserved which of course is the primary objective.

Long-Term Capital Management - Wikipedia, the free encyclopedia

None of the above should be taken as criticism of your work...but perhaps there will be something there that provokes further thought.

 

[Yamato]

With some parts I agree, and some parts I didn't understand. But to make it quick, I would reply about this: "When Whole Market / Fundamental Re-alignment moves are occurring their should either be a circuit breaker in place to stop all trading or an equalization mechanism that kicks in which in effect neutralizes the drawdown beyond an acceptable level".

My idea is that:

1) this is a discretionary intervention and it cannot be done
2) we are unable to identify when such an event happens (if there were a way to univocally identify such an event, then it could be coded, and if it can be coded, then it can be done).

If what you're suggesting is a rule such as "when there's a high volatility, do not trade", then I can already tell you that this reduces the profit, and we cannot create a rule that decreases the profit (it decreases the sharpe ratio and all other metrics as well).

 

[wprins]

With some parts I agree, and some parts I didn't understand. But to make it quick, I would reply about this: "When Whole Market / Fundamental Re-alignment moves are occurring their should either be a circuit breaker in place to stop all trading or an equalization mechanism that kicks in which in effect neutralizes the drawdown beyond an acceptable level".

My idea is that:

1) this is a discretionary intervention and it cannot be done
2) we are unable to identify when such an event happens (if there were a way to univocally identify such an event, then it could be coded, and if it can be coded, then it can be done).

If what you're suggesting is a rule such as "when there's a high volatility, do not trade", then I can already tell you that this reduces the profit, and we cannot create a rule that decreases the profit (it decreases the sharpe ratio and all other metrics as well).

Hello Travis,

Firstly again my sympathy with the recent turn of events, although from where I'm sitting it's not entirely without positives.

Secondly allow me to jump in re the "circuit breaker" idea, and simply add that if I understand the suggestion then I'd suggest that it would not fall under discretionary intervention since it should be implemented as an automatable explicit rule that would act as a circuit breaker. Indeed, having an "uncle point" is actually already such a crude "circuit breaker" rule, albeit simply utilizing a fixed level as circuit breaker.

In my own model development work I've recently been pondering/working on several ideas in this regard, but suffice it to say that one (perhaps obvious, but I don't see why it wouldn't work) idea is that one can calculate a yield curve for the combined model/system comprised of the multiple systems in play. Then, this yield curve, which should be a straight line on the log scale, and which should have a very high goodness of fit coefficient, can then be adjusted downwards so that it runs just under the actual model's equity curve.

All else being equal (meaning, the system continuing to work as expected) you then effectively have a kind of "trailing uncle point" on your entire system equity curve, which if it gets reached is a strong indication that somethings not right. Additionally to this another idea I've played with is some dynamic position sizing strategies that dynamically and gradually adjust the overall positions being traded based on recentime system performance. This also appears pretty promising.

Anway, here's an image to help explain what I'm getting at. Imagine the trendline (yield curve) below, adjusted downwards so it runs under the actual yield cuve and using this as your circuit breaker. Both images are of the same model, one is log scale, the other isn't. The log scale one is arguably more relevant.

 

[Liquid validity]

If what you're suggesting is a rule such as "when there's a high volatility, do not trade", then I can already tell you that this reduces the profit, and we cannot create a rule that decreases the profit (it decreases the sharpe ratio and all other metrics as well).

I have to say I agree with the volatility observation:
http://www.trade2win.com/boards/att...nals/122650d1316862292-my-journal-2-snap4.gif
http://www.trade2win.com/boards/att...nals/122606d1316767449-my-journal-2-snap1.gif
CBOE.com

The drawdown has happened at a time of increased volatity.
I can see what you are saying about the affect on profit.
Personally I would rather focus on risk and drawdown as the primary factors,
with profit as a secondary factor.

Why are people using sharpe ratio, sortino ratio and all that crap? With such unreliable subjects you should not give yourself a false impression of reliability by using all these metrics, that just paralyze you.

You should just use profit as a metric. How much money did such system make in the forward-testing? **** everything else.

**** margin.

**** Sortino.

I agree with that completely, LTCM had a high sharpe which was misleading.
The links below highlight some of the flaws of measuring risk with one number:
Spotlight on the Sharpe Ratio: Part I - Seeking Alpha
Spotlight on the Sharpe Ratio: Part II - Seeking Alpha
Spotlight on the Sharpe Ratio: Part III - Seeking Alpha

Just my personal thoughts, obviously how you progress from here is
your choice and yours alone.
Good to know your backers haven't given up yet.
Waiting until conditions are more favourable is a good move.

 

[brettus]

With some parts I agree, and some parts I didn't understand. But to make it quick, I would reply about this: "When Whole Market / Fundamental Re-alignment moves are occurring their should either be a circuit breaker in place to stop all trading or an equalization mechanism that kicks in which in effect neutralizes the drawdown beyond an acceptable level".

My idea is that:

1) this is a discretionary intervention and it cannot be done
2) we are unable to identify when such an event happens (if there were a way to univocally identify such an event, then it could be coded, and if it can be coded, then it can be done).

If what you're suggesting is a rule such as "when there's a high volatility, do not trade", then I can already tell you that this reduces the profit, and we cannot create a rule that decreases the profit (it decreases the sharpe ratio and all other metrics as well).

Travis.

Can i ask a simple question? In regard to the losses in silver for example, was that just one trade that made up the 10k loss? Does it average in? Or did it buy more? I'm slightly confused how a system can have 37k? drawdown and drop 24k in a day. Also it seemed like it was making far too much money in the beginning for an equity of 160k. I believe 20% a year is an adequate return on a mechanical system in a year. Proving that your system has longevity has a much greater bearing than return. I believe you have learn't this now though.

 

[Yamato]

Hello Travis,

Firstly again my sympathy with the recent turn of events, although from where I'm sitting it's not entirely without positives.

Secondly allow me to jump in re the "circuit breaker" idea, and simply add that if I understand the suggestion then I'd suggest that it would not fall under discretionary intervention since it should be implemented as an automatable explicit rule that would act as a circuit breaker. Indeed, having an "uncle point" is actually already such a crude "circuit breaker" rule, albeit simply utilizing a fixed level as circuit breaker.

In my own model development work I've recently been pondering/working on several ideas in this regard, but suffice it to say that one (perhaps obvious, but I don't see why it wouldn't work) idea is that one can calculate a yield curve for the combined model/system comprised of the multiple systems in play. Then, this yield curve, which should be a straight line on the log scale, and which should have a very high goodness of fit coefficient, can then be adjusted downwards so that it runs just under the actual model's equity curve.

All else being equal (meaning, the system continuing to work as expected) you then effectively have a kind of "trailing uncle point" on your entire system equity curve, which if it gets reached is a strong indication that somethings not right. Additionally to this another idea I've played with is some dynamic position sizing strategies that dynamically and gradually adjust the overall positions being traded based on recentime system performance. This also appears pretty promising.

Anway, here's an image to help explain what I'm getting at. Imagine the trendline (yield curve) below, adjusted downwards so it runs under the actual yield cuve and using this as your circuit breaker. Both images are of the same model, one is log scale, the other isn't. The log scale one is arguably more relevant.

Yes, I've talked about this type of idea with many people proposing it before, but it doesn't fit with any of my automation process. Who says that if we fall for a few days, there's a downtrend starting? Is there any backtesting for this? Unlikely. This is all discretionary to me.

Thanks for your efforts and the great detail of your post, but I've reasoned on this before, several times, given the several people telling me to do it, and it never made any sense to me. There's 120 systems, and they're to be considered separately, and we can't put them all together, and assume that if they start losing as a whole, some kind of downtrend has begun, which can be avoided. Same applies to reducing the investment: if you do that, you might very well end up with investing more on the losing days and investing less on the winning days. This whole reasoning doesn't fit with my scientific methodology and smells like discretionary tampering.

Thanks for the detailed feedback though.

 

[Yamato]

I have to say I agree with the volatility observation:
http://www.trade2win.com/boards/att...nals/122650d1316862292-my-journal-2-snap4.gif
http://www.trade2win.com/boards/att...nals/122606d1316767449-my-journal-2-snap1.gif
CBOE.com

The drawdown has happened at a time of increased volatity.
I can see what you are saying about the affect on profit.
Personally I would rather focus on risk and drawdown as the primary factors,
with profit as a secondary factor.



I agree with that completely, LTCM had a high sharpe which was misleading.
The links below highlight some of the flaws of measuring risk with one number:
Spotlight on the Sharpe Ratio: Part I - Seeking Alpha
Spotlight on the Sharpe Ratio: Part II - Seeking Alpha
Spotlight on the Sharpe Ratio: Part III - Seeking Alpha

Just my personal thoughts, obviously how you progress from here is
your choice and yours alone.
Good to know your backers haven't given up yet.
Waiting until conditions are more favourable is a good move.

I don't know if that is what he was saying, that avoiding volatility is a good thing, but I don't see how i could make that work for us. There's no way to back-test and measure how doing such a thing would have saved us today or in the past. How can we exclude that avoiding volatility would not have kept us from making the 40k we made? There's evidence suggesting that throughout my systems' back-testing. I cannot adopt any approach or rule that is not thoroughly back-tested and whose implications i do not thoroughly understand. And this is what you are all suggesting lately. So I cannot agree with not trading with volatility, I cannot agree with using the average of the equity line, and I would not agree with the investors stopping the trading "until times are more favorable" and similar. They have stopped because they considered the systems dangerous and potentially failed. They might not even resume trading again. I think there's a 50% chance that they will not do it.

In the meanwhile, the other long time friend is also not interested any more, or rather: he said he has to think about it. I knew it. When you don't need people, everyone is offering help. When you need them, they all disappear or "have to think about it".

He kept telling me "why didn't you trade with me instead of them?" for a year. And now that I am telling him that the time is right to trade with me, he says he has doubts. This is hilarious.

 

[Yamato]

Travis.

Can i ask a simple question? In regard to the losses in silver for example, was that just one trade that made up the 10k loss? Does it average in? Or did it buy more? I'm slightly confused how a system can have 37k? drawdown and drop 24k in a day. Also it seemed like it was making far too much money in the beginning for an equity of 160k. I believe 20% a year is an adequate return on a mechanical system in a year. Proving that your system has longevity has a much greater bearing than return. I believe you have learn't this now though.

Yes, that was 1 trade that made the 10k dollars loss, as it was one trade that made the 10k dollar win a couple of days before.

The 37k is the equity line of all systems being traded: it represents their profit throughout a year and a half of trading.

Yes, there was a combined loss of over 20k in one day. Well, that equity line has one limit, as pointed out by bbmac. It represents the profit by the systems, regardless of investment. In the first 8 months, only 15k was invested as margin. Then three months with 30k, then a few months with 45k, then 160k, and that's the period when you see the sharp drop, from 37k to below zero.

Another limit is that it represents the profit on period when we were trading 10 different combinations of systems.

Longevity... given that the money invested was different, and that the systems were different, this cannot be said about systems. Nothing negative can be said about my systems based on the equity line. All that can be said against that equity line and what it represents is that there was:

1) either very bad luck (you could be doing everything right, and still blow out your account: there's no 100% risk-free investment)

2) or a mistake in money management


That is not the capital plus profit equity line. That is not the equity line of an investment in stocks. That is merely the profit equity line of a bunch of different systems, requiring different margin in different periods. There was no better way of representing, with one line, what we were doing. Drawing more lines was not possible, because I am already burned out enough as it is.

 

[bbmac]

'....He kept telling me "why didn't you trade with me instead of them?" for a year. And now that I am telling him that the time is right to trade with me, he says he has doubts...'

Wake up and smell the coffee, this is exactly the point I was making in my post yesterday - the one you got upset about. It is always harder to attract capital or any other kind of interest the more unscuccessful attempts you have behind you...it's ironic really as potentially you are in the best position re experience/knowledge you have been in given the experience (s)

G/L

 

[DionysusToast]

Travis

I know you are down right now.

But you have come a long, long way.

Listening to negativity on here about what you are doing won't help.

Only you have your experience. Somebody doing something totally different isn't in a good place to critique. Still - it's a very good way to make yourself look clever.

Like you said - this ratio & that ratio - well it's all the stuff of TA books. Who gives a fook about Sharpe? I tell you what - I know a few full time traders and if I asked them what their Sharpe ratio was, they probably ask me what on earth it was.

If you think you are close to getting this right - then go with that and stay on the path. You have already proven something. How many other guys here could write systems, attract financing and do the other stuff you did? Tell you what - the guy calling you a ginger headed kid couldn't.

This is the path you have walked. Don't deviate because of some dicks on an internet forum trying to make themselves look smart @ your expense.

Pete

 

[bbmac]

Oh what a cynical view of the world the poster above has !

I have read nothing but well meaning and genuine advice from posters on this thread. On an open forum and particularly with the style/content that Travis has conducted this thread it is perhaps inevitable that such comments/advice will come forward - but as far as I can tell such posts are all well meaning and well considerered and a lot of posters put considerable effort into the arguments/suggestions they put forward - often drawing on personal experience.

Travis will get a lot of advice from all quarters - it's up to him what he does with it (or not.)

There is only a dick posting on this thread !

G/L

 

[Yamato]

'....He kept telling me "why didn't you trade with me instead of them?" for a year. And now that I am telling him that the time is right to trade with me, he says he has doubts...'

Wake up and smell the coffee, this is exactly the point I was making in my post yesterday - the one you got upset about. It is always harder to attract capital or any other kind of interest the more unscuccessful attempts you have behind you...it's ironic really as potentially you are in the best position re experience/knowledge you have been in given the experience (s)

G/L

Mmh, I can almost take this tone (certainly not "wake up and smell the coffee", but the rest of your post is all right). But not the tone you used yesterday. Unlike what you said yesterday, I do not keep failing and saying "but now I know...". This makes sense. It's objective and not too insulting (except for "wake up and smell the coffee").

Despite your post being reasonable and the tone almost acceptable, I still do not agree.

This is not an "unsuccessful attempt". There's mistakes in everyone's trading experience, and so there's been mistakes during this year and a half. We cannot say "you failed, period". There has been mistakes and the drawdown has exceeded our expectations and the investors have quit (not even permanently yet, because they're still evaluating). The systems have not failed by any means. Just because many systems had their drawdown at the same time, this doesn't mean that I failed or that the systems have failed.

The capital invested was 160k, profit from a year was lost, plus another 11k. The first year was profitable, the second ongoing year recorded a loss of less than 10% on capital... what I see is once again your constant effort to discredit me.

Yes, the long time friend of mine is less likely to invest, so is my father, but that doesn't mean anything. It doesn't mean they're right. It just means I am surrounded by more idiots than I thought. It doesn't take anything to bet on someone who has already won, and maybe he will not win again in the future. What takes intelligence is investing on someone who has won yet. And that is what I am.

 

[wprins]

Yes, I've talked about this type of idea with many people proposing it before, but it doesn't fit with any of my automation process. Who says that if we fall for a few days, there's a downtrend starting? Is there any backtesting for this? Unlikely. This is all discretionary to me.

Nobody says there's neccesarily a downtrend starting when your equity curve breaks below the yield cuve (although there might be.) But (the point if you will, is this): if your backtests and forward tests over all available data has shown you that the equity curve should *always* stay above (for example) an adjusted yield curve as above, and it then unexpectedly breaks below it (and stays below it) then at the very least, amber lights should come on -- it implies something is, with a high probability, broken. (Whether temporarily or permanently is at that point not known and in any case besides the point. )

It is, in a word, unprecedented and therefore mandates careful scrutiny as to what's really going on, and taking whatever emergency actions was determined in advance to protect your capital, even if that action is simply to say the systems go back onto paper trading/monitoring for the time being.) To emphasise: The yield curve is after all an objectively defined model of what your system(s) behaviour is and should be with an extremely high probability, and any significant deviation from this expectation is therefore by definition not subjective/discretionary and a strong signal that something's wrong. What you *do* in this situation is a design decision of course, but no more discretionary than you picking parameters in advance for the rest of your systems parameters (market, time, levels, moving average periods/types, whatever) and then backtesting these choices to confirm the model "works".

As an aside, this is one of the unavoidable problem with mechanical systems IMHO -- you don't generally really know whether what you've found/captured in a system is due to sheer luck/randomness and/or indeed what the real nature of the edge that you've captured in a given system *really* is. Consequently, one also will in general not be able to know when the edge you have or had may be be disappearing or may have disappeared. However, my point is that such a loss of an edge (or temporary market behaviour alteration, or extreme random exception) will (obviously) show up in your equity curve (as it has in your case) so it's logical that defining some objective rules based on the behaviour of your equity curve is one way to monitor whether your system(s) as a whole is continuing to work and are continuing to exploit whatever edges they capture.

Lastly, I don't see why you say this doesn't fit in with any kind of automation process. The steps to automating this would be something as follows:
1.) Backtest (and forward test etc) your system or systems, and determine the actual equity curve data points for as much data as you have.
2.) Now calculate best fit (R^2) model for the equity curve dataset.
3.) Optimise/adjust this yield curve model so that it falls below your known equity curve for all data points.
4.) Incorporate this into your automated systems, so that they/it automatically calculates the expected minimum yield at every bar/data point in your system alongside all the other variables being calculated/monitored.
5.) On an automated/ongoing basis, compare the current value with the expected minimum according to your yield model.
6.) When the equity value falls below the model, automatically trigger the circuit breaker and stop trading the system (or all systems.)

So, once implemeted I submitthere's no discretion at all in the above, it's all automatic and objective. The only discretion involved at all, if you want to call it that, is deciding exactly how far to adjust the yield curve while setting all this up, and what the nature of the circuit breaker should be. But I would suggest this is no different the many other decisions made about other aspects of systems during the design and optimisation of those systems.

How's the systems performing currently? (The current market realities are probably likely to kill/hurt many systems -- silver's recent moves for example has probably hurt an huge number of traders...)