The opinions expressed in these forums do not represent those of C2, and any discussion of profit/loss
is not indicative of future performance or success.
There is a substantial risk of loss in trading. You should therefore carefully consider
whether such trading is suitable for you in light of your financial condition. You should read,
understand, and consider the Risk Disclosure Statement that is provided by your broker
before you consider trading. Most people who trade lose money.
I know there are other threads like this out there, but I wanted to add my 2c to the topic.
Now that I’ve got about 1.5 years of active experience on C2 and I’ve been both watching a variety of strategies and also trading a few with real money, I’ve noticed some recurring themes, at least as far as avoiding risky strategies is concerned.
In particular, strategies that have blown up that I had some money invested in included Futures Wealth Builder, Volatility ETF Trader, and A Strategy for YM. Fortunately I was sized small and/or exited early on these so I wasn’t burned too badly by these failures.
Strategies that I’ve been following that have blown up are too numerous to mention, but a recent example was AlgoFolio Short.
Common themes amongst all these strategies:
Lack of fixed stops or hedges, especially dangerous on leveraged instruments (futures) and instruments with tail risk (volatility strategies).
Martingale-like trade entry (textbook example: A Strategy for YM)
Any mention of overly complicated signal generation engines like A.I., neural networks, multi-factor models with 15 variables, etc.
Regular large single-trade drawdowns. I’ll arbitrarily define that as >=5%.
It seems like any combination of 2 or 3 of the above will inevitably lead to system failure. All it takes is one really, really bad day.
Why’d you have to go and piss in my Cheerios, buddy? I’m just trying to contribute something. I’ve subbed to and followed many systems over the past 18 months on C2. In this post I’m just focusing on the subset that failed catastrophically.
Please add another characteristics for dust models:
High return in very short period
Many models that show extremely great return such as 30% in a month with only 1 month track records n post in the forum to show how great is their strategy, it is a red flag.
Superhero or Wonderwoman by doing over leverage. Totally agree with Algosystem. Anyone remember SuperBlue aka Blue Group Investment? That’s what I’m referring.
Always make a new system or inconsistent performance and come into C2 with a new name.
Most model that went bust are future models or forex. Due to the leverage you get at c2. Especially forex they can martingale Atleast 8 times. Look at zip4x or diamonds and others. They go bust and recreate a new one after a yr or 2 when a streak of trend go against them.
But no one with real money trades this way. But when u trade with paper money or demo accounts who cares right! Their main focus is subscriber fee, so they try to get attention from the forums and spam pm. Which many new comers or novice see the +xxxx% annual return they can sucked into subscribing. When it’s too good to be true it’s too good to be true. Ask you self if a trader make 200% gains in month why would he be at c2 trying to make $299 off you. If he is so successful why would he need to spam the forums for attention.
But back to the main pt, if c2 has some sort of leverage limit or contract limit on futures or forex it would prevent most of strategy from going bust. A 25k account able to buy $2-3 million worth of future contracts are just not realistic.
Going off topic…yes @GSPTrader, trading is very hard! This is why only a few survive and make money. There is that old saying that ‘90-95% traders lose’ is very accurate.
The only advice I can give you is to try and stay focused and not over leverage. Don’t revenge trade. If your system has a statistical edge in winning then try to maintain focus.
Is there any way to pull data out of C2 on all systems that ever existed here and start to do some analysis on how many have failed, what is the average lifespan, some of the common characteristics of the ones that failed, etc.?
It would be good to be able to look at some data and more precisely quantify what specific markers one could use to predict chance of failure. In the long run, this would be a huge service to C2 subscribers. I wonder if C2 management does this internally? Everything we discuss here on the boards is based on personal experience and observation, in other words, it’s anecdotal. The only study I’ve seen where someone attempted to look at this was last year after the Feb 5th crash when one forum member posted a study about how systems with high win rates and low win:loss ratios suffered failure far more frequently than systems that didn’t exhibit those characteristics.
I am a generation Y and relatively new to futures trading. My strength lies in probability and statistics as well as pattern recognition. In the 2 weeks I have analyzed the leader board, I have drawn conclusions.
I see the list of new strategies are the highest annual return percentages. The subscribers jump on board without any real track record for the strategy.
What jumps off the page is total trade times. You see a majority of the successful trades all falling in a close range. For scalpers, say this would be minutes. All of a sudden, a trade time pops up at 4 hours. You know already that the system produces good trades in minutes. The volatility of the ES makes up plenty of ground, quickly. A poor trade can be erased in a day or two. If all the wins were within certain time limits, and all the losers grouped by time, then you may conclude that the operator is following guidelines.
I recently saw a 300 a month strategy getting heavy traffic based on a few days. I researched and found what I think is a good trader but not profitable. The times to hold are getting greater which to me means that he is better at manipulating C2 stats than his personal margin account.
Experiment yourself. With a sim account of 100,000.00, purchase 1 contract ESP, enter long, and set profit for 1 point or $50 and zero stop loss. Since the market goes up inevitably, even if it is inflated dollars, you will be a 100% winner. Buffet only sells when he is a winner.
There is a live strategy that is getting traffic and the 1 contract wager around XMAS lost nearly 160 points at its lowest. One contract was down 8k and recovered. The trade won 3 points and went into the plus column for all stats. That 8k in potential losses is only visible in the time stamp. I noticed immediately as minutes went to a week, until it was profitable. If the ES would have broke 2340 on XMAS eve, the account would have blown up. This info is reflected nowhere in the leaders stats.
@ESPeonage, it would be good if you can apply statistics to your strategies unlike most C2 strategies which for the most part use human intuition. This human intuition always leads to problems in over leveraging and rule breaking.
Hopefully you can bring in strategies that will have a statistical advantage rather than the traditional ones currently deployed here at C2.
After reading through it, looks like it’s not possible to get data on systems that are shut down or have gone private. This would make any kind of attempt at analyzing failed strategies difficult, I would think.
Hi @PhilD1, if its not possible to look into the past maybe you can start a new analytic going forwards on some current and new system that are published and gather the data for analysis after a say a half a year?
Its a lot of work but it would yield a possible interesting find…lol
While I don’t have a clean-and-ready-for-ML data set handy (although I will certainly think about how to provide such a thing), a good place to start is the Portfolio Time Machine:
It’s still part of “C2 Labs” …which is shorthand for: it’s wonky, not-quite-ready for prime time, but still pretty interesting if you have a technical bent, and want to slog through it.
Look interesting (to the extent that it works - one of the runs I tried just hung) but I think it’s of limited use to what we’re looking for.
The filter criteria are OK but I think most of the ones that are of interest to us are a little unusual, for example, leverage (can be defined however you like), intra-trade drawdown, use of fixed stops, Martingale-like position sizing, underlying signal logic (categorization), % of time in market.
These criteria lean more towards the a priori system characteristic side of things, rather than being ex-post performance stats. Stated another way, performance stats are backward looking. Characteristics inherent to how a system trades are usually forward looking.
Actually, martingale strategies are the bulwark of institutional trading. Having been one myself for over 25 years, I can attest to how effective they can be if you properly manage your position timing and risk. Top-ticking and bottom-ticking any market is a losers game, but using a firm strategy to implement right-sized trades in a scaled-in position can work well if you’re properly trained, well disciplined and have a firm set of rules.
OK, but I’d like to provide stats that seem broadly useful to investors on C2. So if you have some interesting methodologies / statistics that you’d like to see baked into the platform, reach out to me. Maybe I can build the stats / analysis methods into C2 (and even possibly into the Portfolio Time Machine data set). It all depends on how broadly applicable they are, and how computationally intensive they would be to create.
Yes @RexRacer, institutions use those types of strategies combined with hedging in case their martingale starts going out of control. But then they have pretty much unlimited funds relative to the typical investor and the trader must be extremely well disciplined.
Also, the newbie traders tend to let their emotions ruin the trade and hold on too long sacrificing more than half their account or more on a single trade.
A Martingale scale-in method on equities, unleveraged, is usually referred to as value investing.
In futures and forex trading, it’s the leverage and the much-wider-than-anticipated probability distribution of price moves that will kill you. This assumes you’re not using a fixed stop.
Leverage: for each trade calculate the notional value of the position when it was at its maximum size, and divide by the model account equity at the start of the trade.
Intra-trade drawdown: you already calculate these for each trade. Maybe give an average and a st. dev. for the history of trades for a system as well
Use of fixed stops: when trades are entered, can C2 see if corresponding exit stop orders accompany each entry order? If not, then maybe just allow a trade leader to self-report on the system whether or not it uses fixed stops.
Martingale-like position sizing: done as a yes/no tag. If a strategy repeatedly (at least 2 more times) adds more size to an existing position that is in drawdown at the time of adding more volume, tag it.
Underlying signal logic (categorization): other than having trade leaders self-report, for example in the same way they categorize their strategies (momentum, etc.), I’m not sure how this would be implemented
% of time in market: take the average trade duration (C2 already has this data) and divide into the total time that the market that the system trades has been open, since the system started trading on C2.
How might this be used in practice? If I saw a system tagged as using Martingale-like sizing in combination with some leverage multiple above X, and maybe the st dev of the intra-trade drawdowns is wide compared to its average (occasional big drawdowns), I would scratch it off my list of potential strategies without needing to dig further. My gut instinct is that this type of strategy would be very risky.
