We are proposing here some rules that could avoid problems of unknown risk.
The rules should increase the prestige of C2, defining reasonably the risk level of every system.
All what we are writing is an opinion. Nevertheless, the whole stuff is currently well functioning on the organizations managed by our firm: and customers are really satisfied.
POSSIBLE LOSSES ARE, IN THIS CONTEXT, REASONABLY ACCEPTED.
In a trading system:
You cannot insert a trade if there is no stop loss. If stop loss is not included, simply the operation will not be open by the C2 computer. In this way stop losses are forced to exist from the start. This solves many problems also if there is an interruption in the Internet connections, where operations without stop loss can be exposed at no-limit risks.
Stop losses can be expressed in percent manner also: this is mandatory when the trade opens on opening or at a pre-defined time and we cannot know the exact price.
I.e: buy on opening (or at a certain delayed time) the Sp future, stop loss 0.38%.
The C2 system will convert the percent stop losses in prices.
Stop losses cannot be deleted, and can be changed only as follows.
Suppose that you buy SP at 1000 and stop loss results in 1000-0.38% = 996.20 level of exiting.
You can change stop loss without closing the operation, but such a new stop loss cannot be lower than the original 996.20. So you can perhaps execute averages in losses also, but in respect of the 996.20 boundary. Probably no one will mediate on losses.
This is true till the operation is closed.
At a meta-level (on the edit/creating system sheet), the system has a Declared Maximum Stop Loss (acronym DMSL). Suppose 0.5%.
Every new operation opened, CANNOT HAVE A STOP LOSS greater than the declared maximum stop loss DMSL.
In the preceeding example, stop loss is accepted because 0.38% is less than 0.5%.
ON THE FOLLOWING DAYS YOU CAN CHANGE THE STOP, BUT NEVER LOWER THAN 996.20: IF THE CASE, THE CHANGEMENT IS NOT ACCEPTED.
Before that the 996.20 level can be deleted, you should insert before an higher level (say 997) and then, and only then, the 996.20 level can be deleted.
We realize that this is a strong discipline, but money should not be at risk beyond the consciousness of the customer.
I think that a system at known risk can surely ask the customers higher fees and C2 will benefit also.
In order that all the matter will function, there should be also an Historical Declared Maximum Stop Loss, which is the higher one used across the time by a system (acronym HDMSL). HDMSL is used on calculating the EG Risk Index here below.
HDMSL in a system context can only increase and never can decrease.
Declared Maximum Stop Loss DMSL can be changed on the higher side only when every operation is closed. Moreover, no operation can be opened for n days (say two or three) after the Declared Maximum Stop Loss DMSL has been changed.
As told before, HDMSL changes only when DMSL is increasing, reflecting the higher DMSL ever used on the underlying system.
We can at this point create a very useful instrument, what we call the EG Reward Index.
Where EG = [Yearly compounded percent performance] divided by [HDMSL]
That is EG = yr% /x HDMSL
Suppose that a system at known risk has a performance yearly compounded (perhaps after ten weeks at least) as 20% and that HDSML is 0.5%
The EG Reward Index will be 40.
Suppose that another system at known risk has a performance yearly compounded (perhaps after ten weeks at least) as 150% and that HDSML is 5%
The EG Reward Index will be 30, less than the previous exemplified.
The second system should perform at least 200% yearly to have the same EG Reward Index
So 20% can be rated better than 150%…
This is not to say that the world will change, but a customer knows that one system risks 0.5% stop loss while another risks 5% stop loss. It can decide better on its own.
The higher the EG Reward Index, the lower the risk of surprises.
Systems that do not follow the rules should be categoricized as SYSTEMS AT UNKNOWN RISK and obviously cannot have the EG Reward Index.
Or conventionally an EG Reward Index as -100.
Obviously every one can prefer to use the Systems at Unknown Risk that perhaps can have very high gains.
After that, Sharpe ratio can also be used. But currently we can have high Sharpe ratio with systems at unknown risk, BECAUSE ALL OPERATIONS ARE ACCEPTED WITHOUT A PRIORI CONTROL. IN THIS WAY, EVERY RISK IS EVALUATED A POSTERIORI AND NO ONE CAN AVOID THE LOSS OF THE FULL CAPITAL. This can succeed obviously also with systems at known risk, but at least we are trying to exercise a certain control. The current situation can stimulate (on providers) a worst competition, if they want to be well classified.
Obviously, what written here cannot cover all the matter: when a rule is given, there is always someone who finds a way to avoid the rules in some unexpected and innovative manner. This was said by Cicero more than two thousands years ago.
Nevertheless, I think that the exposed rules can be useful for many reasons. Because this is an opinion and not the Holy Grail, obviously it can be modified (as well than discarded). We repeat however that this is currently well functioning.
We see in all the exposed matter that the customers can be more confident.
If the customer is using auto-trading, he or she can know something more about what can happen to his/her money.
P.S. Currently, some of the greater brokers do not have (or do not permit?) contextual stop losses. C2 could be the one that can start a little revolution. Please consider the following:
Sell short 10 contracts of SP. Wait for the execution and after that insert the stop loss. Well, when the execution is already done and before that the stop loss in inserted, the Internet breaks itself and while you are short the market begins to go up, up, up
this is a disaster and can be the ruine.
They say: you can use the telephone! But this is the moment in which all phone lines are occupied
believe me, this happened and not only in october 1989 but alto in september 2001.
From this, it is clear that the contextual stop loss is almost mandatory. So there is this reason also to change our philosophy.
In our firm, we are currently using some more restricting rules, for example no more than n loosing operations can be opened in a day, or a system cannot suffer more than (for example) 2% losses in a day. After that, for the underlying day, every operation is closed.
The 2% should be declared on the system sheet and should not increased but only decreased.
Vox clamantis in deserto.
Ernesto Giorgi seleukos
"You cannot insert a trade if there is no stop loss."
Some advisors have trades that last for 15 minutes or 15 hours or 15 days or 80 days. They might be in the market 100% of the time, and switch from long to short. They might put out a signal that is good for QQQQs or Rydex Nasdaq or emini Nasdaqs or options on Nasdaqs - these instruments generally move similarly, but haeve different leveraging and are priced differently.
Some traders will receive signals that are designed for Russell 2000, and use it to trade S&Ps. Again, pricing very different. Some traders have their own idea of money management or stop losses, and it might be considerably tighter or looser than the advisor’s. Especially if they autotrade, and cannot monitor the trade all the time.
How do you “enforce a stop loss”?
As written, new problems can arise. Nevertheless these systems can be categoricized as sytem with no pre-defined stop loss, or, as told, at unknown risk level. This is surely true. Nevertheless, no one says that such systems should be eliminated. The target is that the customers should be well informed of the possible stop losses. No more. In our organization we eliminated all systems that can not have a predefined stop loss.
Last example is our Energy system that will reverse. Now we are changing it, because we imposed a stop loss as 1% when reverse. But what if a second reverse in a day?.
Substantially we are working to lower potential risks.
Some aggressive techniques should be discarded, in particular when using the money of other persons.
This is not the Decalog. In every case to fivide systems in some categories could be useful for the new customer, not always skilled as you can be.
I do appreciate your effort to push things in a more customer-friendly direction, though
If I understand you correctly, we would then have two kinds of systems:
A. Systems that always use stop loss orders and that specify a maximum stop loss in advance and that are not allowed to exceed it in any trade
B. All other systems.
I think this is a great idea. I have only a question whether this would be allowed legally, since C2 would interfere with the advice being given. If that would be a problem (I don’t know) then perhaps it could have a more passive form, where (1) C2 gives some kind of ‘stop loss certificate’ if the trader adheres to his own stop loss rules, specified in advance, and (2) C2 gives some kind of warning to the system vendor if he tries to enter an order without the accompanying stop loss order.
Jules
Very good idea!! A stop order with every new trade.
Especially because I am already doing this in all my 3 systems now. This eliminates some problems and is very helpful for subscribers.
(If you do not place a stoploss with your opening order it easily happens that you misplace or mistype your stop order later, which happened once in my new futures system
- vicinoo! trading - Futures (beta) +258% - and resulted in one huge loss. Nonetheless it was my fault… But now cannot happen again. Unreachable C2 servers etc. are other problems that you mention and that should be considered very serious by subscribers when choosing no stop-systems.)
I hope with this feature subscribers will gain more valuable insights and can better identify those no stops-high win ratio-undefined possible drawdown-systems (of course systems this way can work…sometimes) and those systems that take losses fast and have a lower win ratio but control risk.
2 problems with the idea:
1. Even if everyone agrees with the idea of obligatory stop loss orders, how do they agree on the level of the stop loss?? For one subscriber 5% SL is OK, for another 2% is too much…
2. There is nothing to prevent the vendor to enter a new trade in the same direction, once he was stopped out. Thus instead of a 5% loss he can easily get 1-1-1-1-1% losses.
Because of these problems, the idea wouldn’t work.
If your vendor doesn’t use stop losses, there are 2 solutions:
a. Don’t subscribe to it.
b. You can use one on your own.
Simple as it is…
As written, we can introduce some limits to the number of operations. Or, better, as written, A SYSTEM CANNOT LOSE MORE THAN ONE FIXED PERCENT IN A DAY.
Please re-read the original document: I believe that there is a reply to your questions.
The scope is that the subscriber should be protected. No more. Remember that currebt system with no protection can go on exactly as now. There is no reason to prohibit a new, protected category.
Ernesto Giorgi - seleukos - we are here to invent some idea…
- As I understand it, the vendor will determine the stop loss level. He is the one that (hopefully) understands best what is happening.
2. (Apart from the reply of Seleukos:) Is that realistic? But even if it is, I think that it is comparable to leaving the house with your door open or with your door locked. If someone really wants to break in, the lock can’t prevent it, but the chance that someone will try is much smaller with the door locked. We may add the rule that a system will be expelled from the group if it does such things.
Obviously, any sensible vendor would use a SL. (and I am guilty as charged) If not, any sensible subscriber should/could/have to decide if he is willing to take the risk of not having a SL or not. So to try to interfere with a system and put in an extra safety valve, goes against sensible capitalism. >>> Systems without SL sooner or later will become extinct.
Eventually, no matter how safe you are trying to make C2, subscribers always have to take risk, and a leap of faith, namely, that the system what they subscribed to will keep up the performance as it has shown in the past.
If we want to give Matt more job, sure, we could ask for a Certified Stop Loss User stamp for vendors who classify, or we could just take responsibility for our actions…
"2. There is nothing to prevent the vendor to enter a new trade in the same direction, once he was stopped out. Thus instead of a 5% loss he can easily get 1-1-1-1-1% losses. "
While that is true I suspect most vendors wont do that. Once a trade is stopped out it usually forces a fresh perspective (do I want to get back in?)
Also I think many vendors don’t take their losses exactly because they are trying to uphold some mythical win/loss ratio so as long as they dont accept the loss it isnt booked. So for them to take 5 small losses (which would really hurt their stats) would seem unlikely.
Anyway I agree that a C2 enforced stop for those systems that sign-up for it would be a great idea. Too many vendors just let trades go crazy against them hoping that they will come back - if they do then they are considered brillant, if they don’t and the loss is bad enough then they just walk away from the system.
Simply relying on vendors to keep their word (e.g. state their max drawdown in their system description) does not work. C2 needs to provide a mechanism to enforce both hard stops and maximum positions sizes so clients can be comfortable that vendors will do what they say they will.
I fully underwrite.
But I ask:"How many guys do have the capability to understand that Pete shows?"
Really, not a great lot.
But they are around and we should forecast that someone, after loosing money, will blame C2.
C2, in my opinion, should avoid these situations. Firmly.
It is a game: the game should have the stop loss.
Before October 1989 there were around, say, 13500 guys without stop loss.
They are died, but now there are other 13500 different guys but always without stop loss…
ECCLESIASTES: There is nothing new under the sun.
In october 89, a loss of 29% about in two hours killed many of them.
Moreover if your stop was too large, you were bypassed by the crash.
Only tight stops pre-existing captured at beginning of the crash some ‘white flying’ buyers. After ten minutes, there was the chaos.
We can demonstrate that, where there is a trend, without stop losses we can gain enormously. And lose consequently when the trend will change…
Moreover, for a strange mind process, the guy without stop when gaining believes to be a kind of god. He pretend that market shall (not should) follow him and not viceversa.
Try to gain in currencies now without stop…
Or on the very simple SP…
Well, having a forced stop against a big market meltdown might make more sense. I am curious what Matt’s take on this is.
I understand the need of protection for subscribers, but if you are a vendor, not having forced stop losses can be an advantage for you.
You can advertise your system as one HAVING stop loss orders all the time, thus subscribers looking for such a systems would rather choose yours instead of others. Remember, C2 is a giant competition for vendors…
In Italy, the Fineco group, suspended the activity for the unique danger: the loosing customers screaming.
The danger are not the customers that are loosing.
The danger are the ones that loosed and scream, because they say that they were not informed and so on: when two or three strong systems are going bad, it is a disaster.
You can write on tobacco that is is dangerous…but you are condamned.
I repeat: if a constraint is mandatory is respected. Otherwise the auto-regulation can be waited for two centuries.
I think that C2 should protect itself.
After all: why some of us are asking for a protection? I am a developer and I am not asking the protection for myself…
Two days ago there was a lady from Poland who said: "I have opened an operation with the xy system… is it there a stop for my operation?"
Someone of you probably remember such a post. I think that any comment is useless.
"The danger are not the customers that are loosing.
The danger are the ones that loosed and scream, because they say that they were not informed and so on…"
Claim:
"r2 0.98545 (exceptional) called also good or fitness or linearity, probably more important than the Sharpe ratio. The value 0.98545 means that the equity line is in practice a stright line.
Maximum draw down -5.381% daily computed
Yearly return 41.14%."
Reality:
“Max Drawdown 26.33%”
"Annualized % 2.83% over 101 days"
Yes, I agree. The real danger is when you misinform or make claims which are simply and obviously untrue. This is where C2 needs to be VERY careful.
I can’t see C2 saying every system needs to use stops (even if I think every system should use stops). But C2 should be sure vendors don’t make claims that they will be profitable, will limit risk, will use stops, etc. if those claims are not true or if they can’t be known. Stops do not guarantee limited risk and PAST PERFORMANCE IS NOT NECESSARILY INDICATIVE OF FUTURE TRADING RESULTS.
You are proposing a solution to “avoid problems of unknown risk”. I can agree that better risk measures could be reported for C2 to compare systems. For example, it is not immediately clear how the Sharpe ratio is calculated, and how many datapoints are used. Ideally each Sharpe ratio should be reported with a confidence interval, and I would bet that for a large number of systems offered on C2 these confidence intervals will overlap–so no statistically significant difference in terms of risk-reward (Sharpe ratios reported by Morningstar suffer from the same problem, however).
I don’t agree that stop-losses might help to “avoid problems of unknown risk”. In case of large gaps, bankruptcies and illiquid instruments, stop-losses do not offer protection at all, and might in fact increase risk due to excessive slippage. Remember, when your stop is hit during a free fall, you are at the mercy of your broker, who doesn’t guarantee anything during volatile times. Your position might not be closed at all, or at a very unfavorable level. Everyone who has been stopped out in forex using stops at times unanticipated news came out will understand what I mean. In addition, stop-losses can only limit the risk (if at all) for a single period. What is the risk of a system with a stop-loss of 1%? What is the risk of that same system, knowing that it has a 50% chance to hit the stop-loss for 100 consecutive trades?
The only way to limit risk is to hedge a position. And mandatory stop-losses could actually limit the ability to hedge positions. In your proposal it would be possible that a well-hedged system (without stop-losses) would be displayed as a system with “UNKNOWN RISK”, while in fact its risk is lower than any unhedged system using stop-losses.
I think it is meaningless to talk about trading systems in terms of KNOWN risk. The risk of any system is unknown as long as future returns are unknown, and can be estimated at best in terms of expected risk.
I agree and would like to add the following:
Why not making an option within C2 that automatically sets a fixed stop loss everytime a new trade is opened. This may be typed in the "Edit System page" for example.
I am trading a forex system here with a stop loss that has a fixed 75 pips stop loss. I am well aware that there are more sophisticated stop-loss calculation methods out there (ATR or Stdev stop losses), but what I want for my subscribers is complete transparency.
You can always change the SL if the trades goes in your direction, but not the other way round.
Best regards
Jean Marc
Science trader,
I agree that stop losses won’t take away all risks. But the proposal of Ernesto / Seleukos will make it easier for subscribers to chose the trading style that they like, and the kind of risks that are associated with it (many small losses versus a few great losses). It has happened more than once that a system vendor promised to use a stop loss and then didn’t do it. This is of course very frustrating for subscribers. The proposal will help to avoid this. The proposal as I understand it is not that every system should apply stop loss rules, but that there should be a clear distinction between systems that consistently do that and other systems.
For example, I know at least one system (Trend Toast) that has this trading style of accompanying every open position with a stop loss signal to close. I suppose there are more of such systems. These systems deserve a label. Much like systems are now labeled as ‘futures trading’, or ‘daytrading’, this is also an important label.
In addition, Ernesto suggests that those systems (if they want the label) should be forced to maintain their style. I agree with that too.
Jules
>The proposal as I understand it is not that every system should apply stop loss rules, but that there should be a clear distinction between systems that consistently do that and other systems.
"In a trading system:
You cannot insert a trade if there is no stop loss. If stop loss is not included, simply the operation will not be open by the C2 computer."
Sounds like he wants all C2 systems to have mandatory hard coded stops… I’m I reading it wrong? Also:
"We are proposing here some rules that could avoid problems of unknown risk. The rules should increase the prestige of C2, defining reasonably the risk level of every system."
This is untrue: stops may help limit potential loss, but they cannot “define risk levels”. Making such claims is simple fraud.
I DO agree that systems should be held accountable to their own rules and claims. This could be reflected in the C2 “ratings”. For example
Seleukos has an “excellent” rating. However he claims:
"Maximum draw down -5.381%
Yearly return 41.14%"
C2 shows:
"Max Drawdown 26.33%
Annualized % 2.80% "
We should be correct. And honest. On our example on our Energy system we were speaking about our records without leverage.
In fact the following the one you are reporting says: using a leverage three you could face a draw down as 15%.
Really, we are using a leverage higher than 5, so draw down could be 25 or 30%…
So please be correct or reporting texts. The performance of 41% yearly compounded is in our experience WITHOUT LEVERAGE. We have informed that every one can ask us the full tracks of the past, where you can calculate on your own.
For some days we suspended the ENERGY system because we decided to remove the auto-reversal condition. Too dangerous.
Now it is on the road again.
We are here to repeat: please do not report the text written changing the sense. It is not useful. It leaves in the mouth the strange sensation that there are unfair fellows, with no apparent reason.
We repeat that using leverage one the system had and probably will have a draw down of 5%.
When writing on the computer, the computer does not blame you if what you are writing is not correct.
Probably you have not read well. But it is not the first time and this behavior can induce to desert the forum.
Another thing: we think that energies in this period are very difficult. We would see someone other operating on energies. It is not easy applying a pre-determinated stop-loss.
Another example is when was reported the sentence:" If not stop loss, you cannot open operations." Where conclusion was that this will ruine C2 bla bla. Every one has read that the sentence was related to system that decided to be in the category of stop losses declared and not for all. So the problem does not exist if a developer will stay with the old method.
Another of you says that if the market is crashing stop loss is useless. True, but I wrote that in such a day the stop loss is good if pre-existing. But no one reads… if you put the stop when there is the crash, it is too later. Why this way to argument when reading papers everyone can see that what was written before is different?
To avoid this, the writer should continuously repeat "No! I wrote another thing!" . I find this ridiculous.
The rule probably is: " I can criticize and tell something different from the truth because no one will read the original." This is true. And this is a reason why probably is useless to write in a forum.
