Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
The ALMI actually is becoming a semi-viable trading option. Up until about 3 months ago there was only one market maker and their spread was set at 7points either side of the ALSI, which was huge! Thankfully now there seems to be another market maker and they seem to consistently be the same (or within a point) of the ALSI.
The only problem is that the volume still isn't very high (it's about 1500 deals per day). So what is needed is that you chart the ALSI, but make the trades on the ALMI. You also need to either have mental stops, or alerts when price meets a certain level on the ALSI. This is because the ALMI might not trade until its way past your stop.
Another thing is that the brokerage to trade the ALMI is generally the same as the ALSI, except the ALMI contract size is 1/10th the size. Basically what that means is you need at least a 30point move to overcome your round trip costs. So scalping is out of the question generally.
However other than that, the ALMI is slowly becoming a more viable option. Or at least a good learning ground before moving up to the ALSI.
Yes I do this but I do not use it to turn trading on and off - as you have found switching off may mean missing big moves, (if that is your trading style). I've done tests for strategies that rely on few big winners its always less profitable.
I use it to influence my % risk. The closer my equity curve gets to an MA of my last 20 trades the less risk I put on. I find it helps with keeping me trading live so I can get back in sync, if I'm out of sync.
High frequency strategies that take numerous small trades may be more suited to the equity curve/equity curve MA cross approach. Or if you have a purely mechanical approach where you tweak parameters via Rolling Walk Foward Analyisis, you could use the approach to know when to re-optimise.
And in my book anything that keeps the focus on risk is a good.
I agree with your finding and thanks for sharing them. Equity curve management is just another buzzword in the series of buzzwords invented by talking heads in this area. An example: if the equity drops below it's moving average because of choppy action and then the system misses the trend signal, when the simulated equity goes above the average the system misses half or more of the gains. That is disaster. You either believe in your system or not. This is what it is all about. You can't fix a lemon by squeezing it.
I have not heard mention about utililizing independent strat equity, nothing to do with the strat itself, a out of market profitable or not strat wont trade on a curve break , but other market style strats should pick up when the other leaves off. same stock, multi stat setups specific to our common market conditions, ie; trend up and down, atr range , speed an allotted accounts to trade equity independently, ... 20 great strats do not trade all the time regardless yet can be at the ready as market conditions permit, problems associated with curve is reoptimising the proper curve and when, yet a tight fit strat should hold a tight fit curve verses a 1 strat fits all bit, will be testing multi equity curve in near term, there is a strat picker dll out there the impliments stats bases on curve setting in paper mode and sets it live once equity curve conditions are met, sets it to sleep when broke back into paper mode for re analysis. not rocket science, i curve 20 different charts same strat different settings independenly, i do not automate paper or live account switch, next on my todo list. hope this was some insight beyond the perfect strat to beat the curve, it more like how one curve can help the other. a strat must run live or paper 24 hrs a day , your strat is your servant keep it serving you or thats when you miss the next big move.
Not a vendor, just a trader. I do not sell anything. Current studies involve EasyLanguage, Money Management and Technical Analysis. Do Enjoy: Fibs, S / R's , Eliotts, Cycles, Trends , Skews. Not Enjoy: fundamentals, Main St Media . It's all about keeping your emotions out of trading and work a system. Appreciate all insight as will be returned. Scotty B.
I have been experimenting with an equity curve. I had stated previously never could get an equity curve to benefit my trading. But I recently added some new strategies to my trading mix that are significantly upping the number of trades I take. I am finding an equity curve to be of great value. Some key things at least in regards to my situation. My system does not rely a few big winners for profit. My profits are spread out over a lot of trades so going to SIM mode when below the curve and live when above works well. I think a system that relies on a few big winners could have problems because if you miss that big trade while in SIM can really skew your results. I have never liked a system that relied on a few big winners anyway because there are many things that can cause you to miss that big trade. A system that spreads profits out is much more forgiving imo.
Will give up some profit using a curve but less draw deep draw downs. Even a good system when market conditions change can have sustained draw downs. Does not mean system has gone bad but market conditions just are not right for the system. A curve can allow you to wait out those conditions without destroying your account. There is possibility to get stuck going above and below the curve every other trade which can cause some problems but I have found that is less of a problem than potential for a sustained draw down without a curve. I have also found curve does not work well unless a system takes a good number of trades. A system that takes a small number of trades at least in my experience can actually be hurt by using a curve.
There are some mental hurdles to overcome if using a curve. Main one is it is hard to go to SIM when below the curve. Only way you get above curve is to have some trades win in SIM. So that means you will will be in SIM on some profitable trades. The feeling of dang I missed a winning trade must be dealt with. I deal with it by looking at my overall system and not getting caught placing to much importance on individual trades.
If considering a curve treat it like everything else and run your back test results through it to see how it effects your system as not every system will benefit and it can even hurt.
"The day I became a winning trader was the day it became boring. Daily losses no longer bother me and daily wins no longer excited me. Took years of pain and busting a few accounts before finally got my mind right. I survived the darkness within and now just chillax and let my black box do the work."
There is an article by Thomas Stridsman titled "Martingale Bet Sizing in Drawdowns" covering this approach of increasing the risk of a trendfollowing strategy slightly during drawdowns. (Google will lead you to the pdf.)
For the strategy in the article it helped to shorten the length of the longest drawdown but didn't reduce the deepness of drawdowns.
Platform: TradeLink, OpenQuant, considering anything that works...
Trading: if it trades...
Posts: 94 since Oct 2010
Thanks Given: 24
Thanks Received: 39
Thanks for starting this thread.
As usual I was Googling this and it brought be back to nexusfi.com (formerly BMT) as one of the members had already attacked the problem. Thanks very much.
I have a couple of questions and share some experience / results.
Would those that have implemented equity curve analysis mind sharing the implementation method on NT? Even if it isn’t the code, a bullet point summary of the steps to implement?
As per my post to Mike, isn’t looking for a smooth equity curve just the same as looking for a high Sharpe ratio?
As anyone played with trade dependency (see my post here and this short piece)?
Trade dependency
I did a little bit of research into this a while ago, I was varying the size of risk based on the previous trade. That is, if the previous trade was +ve then bet double the current percentage. If the previous trade is –ve then bet half the current percentage. I didn’t find it improved performance at all.
Daily PnL Filters
I’ve had some success with Daily Max loss filters in backtest. However, implemented in a way that says don’t scaling into any further positions with mean-revision strategies. However, I’ve seen negative results when making the system aggressively close positions when losses are occurring (with mean-revision systems). My theory here is that when we are aggressively closing trades, it is cutting off the possibility of the market reverting and reducing the losses. This isn’t dissimilar findings from Ernie Chan I think (he mentions this in his first book).
When Ernie spoke at this the London Systematic Traders Club he mentioned that his updated approach (to be published in the next book) sounded similar as @mokodo’s where he varies the size of his risk based on the performance of the system, taking risk off aggressively is the strategy starts to “under-perform”.
This equity curve is a back test from 2008 – 2011 of ES with a mean revision system with no daily PnL Filter.
This equity curve is exactly the same system on exactly the same data with a Daily PnL filter that stops the opening of new trades when a threshold is breached.
That is by no means a bad equity curve, especially for a TF system. A few periods of ~breakeven, a couple of minor dips that are recovered from quickly. I'd begin sim trading that live right now to start seeing if live results are reflective or your backtest results.