Risk Reduction = Poor Performance ?

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Risk Reduction = Poor Performance ?

Post by Chris67 » Thu Dec 15, 2011 3:18 am

I remember reading an interesting interview with Paul Mulvaney of Mulvaney Capital Management recently (someone I know quite well also) who made a great case for why trying to reduce volatility and draw downs and pander to investors "wants" can ruin a very good system.

I have several parameters I use which reduce risk and volatility. If i backtest out over , say 20 years, and I have what I think is a very robust system that shows a CAGR of 90% and a max draw down of 50% then I "slip in " a risk reducing equity parametr and I see I get a nice 35% CAGR with a max draw down of , say, 20% - all nice and comfortable and fitting more closely with what I believe are realistic expectations and what I can live with
Its interesting however that taking the above example - that particular system run raw witha normal fixed fractional money manager is up 40% for 2011 and with the risk reducing M/M is down 2.7% on the year
Thsi really does back up what Mulvaney says in many respects

Similarly I notice using other risk reducing parameters such as limiting instruments within groups, etc has had - from what I am seeing , a similar effect this year which really does bring back into the spotlight the conundrum of running a business and/ vrs making returns and the difficult decisions we face

It seems to me that everything you think will work well and reduce risk etc seems in the end to not work anything like as well as letting systems fly

Any comments ?

rabidric
Roundtable Knight
Roundtable Knight
Posts: 243
Joined: Mon Jan 09, 2006 7:45 am

Post by rabidric » Thu Dec 15, 2011 5:13 am

It's pistols at dawn, and you have a choice:

1. A super lightweight ceramic composite 9mm semi-automatic with a memory foam grip that moulds directly to your hand, underslung flashlight, special tungsten tipped ammo, exquisite jewel inlaid carvings on the barrel.

2. A sawn-off shotgun.

Moto moto
Roundtable Knight
Roundtable Knight
Posts: 427
Joined: Mon Jun 01, 2009 4:12 am
Location: once again in the UK

Post by Moto moto » Thu Dec 15, 2011 5:24 am

rabidric wrote:It's pistols at dawn, and you have a choice:

1. A super lightweight ceramic composite 9mm semi-automatic with a memory foam grip that moulds directly to your hand, underslung flashlight, special tungsten tipped ammo, exquisite jewel inlaid carvings on the barrel.

2. A sawn-off shotgun.
depends on what the other guy has!

Chris....do you give the client what they "believe" they want, or do you try and make money.....another trade off.

Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 » Thu Dec 15, 2011 5:57 am

I always try and make money - but I do it in a way that gives it to investors they way that they want it based on what share class they are in - thats why, I guess many TF's have many types of share class ranging from institutional low volatility to high vol double leveraged turbo class ?

I do think however they a lot of investors in traditional low volatility programs are probably making an in-efficient use of capital

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Thu Dec 15, 2011 9:28 am

How many datapoints does it take to convince person X that claim C is true? One?
... taking the above example - that particular system run raw with a normal fixed fractional money manager is up 40% for 2011 and with the risk reducing M/M is down 2.7% on the year This (one data point) really does back up what Mulvaney says in many respects

gunter
Roundtable Fellow
Roundtable Fellow
Posts: 83
Joined: Wed Aug 24, 2005 7:25 am
Location: George Town, Cayman Islands

Post by gunter » Thu Dec 15, 2011 10:57 am

I did quite a bit of backtesting on this to see if I could get an undercapitalised account to become profitable. I reduced size whenever drawdowns hit a certain % or if x % of trades were unprofitable. The effect of these drawdown reducing techniques was that the system would have smaller, but longer drawdowns. Basically, at the start of new big trends the size taken was always too small. In terms of MAR, I believe it actually went down once I implemented these measures.

However, I did not pyramid into new trades and I had a very small universe of instruments (6 to be exact). If you expand the number of instruments traded then perhaps the results could change.

On another note, increasing trade size in the above example also did not work that well. Drawdowns became larger and the open risk on new positions became unacceptably high. However, depending on parameter tweaking / optimisation (i.e. to avoid blow-ups), this got the account out of drawdowns the fastest. Only problem was that I did not and still do not trust those results.

Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 » Thu Dec 15, 2011 11:38 am

Sluggo - You r point is valid... BUT .. if you are running a TF Fund out there in the public view and you had, say 2 years of -2.5 and up 2.5% as your risk reducing techniques were working great but those 2 years presented other TF's with two 30% years then Im afraid you aint got a business any longer

Toosday
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Fri Jun 03, 2011 11:21 am
Location: Austin, TX

Post by Toosday » Thu Dec 15, 2011 6:32 pm

Chris,

When reading your post it sounds to me as if there are two issues. Uncertainty about the hit to returns by decreasing risk and managing your clients expectations.

With regard to the second, I have found that their must be a reconciliation of utility curves (I define utility curves as the weight put on the various statistical moments of the return distribution) between the money manager and the client. This usually involves educating the client and interpreting their goals, which they may not be able to state statistically
.

As far as the first issue, in my experience, when you reduce risk you reduce return. The question is whether you reduce your risk adjusted return. The risk measure you use is an individual thing. I think that if you let yourself be guided by your personal utility curve, the result will become much more comfortable (ie there is no right answer just a reward/risk/skewness/kurtosis trade-off).

Post Reply