annual profit target

marriot · Post by **marriot** » Sun Mar 04, 2012 2:38 pm

Attached is Winton performance.
Looking at 2008 / 2010 i wonder if they have some kind of proifit target whose code is something like "oh my dear, we are up 20 per cent this year, let us exit almost all".
Do you think it is possible code this?

rabidric · Post by **rabidric** » Mon Mar 05, 2012 3:52 am

i see no evidence to support your supposition in that table. Their monthly volatility appears more or less the same all through each year.

I would very much doubt they have an upside target. For them, the only thing that matters is making sure their DDs don't breach 10%, as that is probably the number one priority of the allocators of the institutional client megamoney.

Those clients require TF exposure for one reason or another, but no-one in the allocation chain wants to lose money at any cost tbh. This is the core of institutional money manager thinking: get exposure to xyz in the safest apparent way.

These allocators basically are short puts on their careers. One major bad allocation and they are screwed, but as long as they turn in mediocre upside along with the herd and according to their check-boxes, they earn a good living.

The mega-funds know this and structure their business model accordingly.

AFJ Garner · Post by **AFJ Garner** » Mon Mar 05, 2012 8:40 am

rabidric wrote:
The mega-funds know this and structure their business model accordingly.

And for those who want to make real money instead of tinkering around on the margins like most of us on this forum (myself included) how very sensible to give the customer what they require. Is it not said in the retail world that the customer is always right? I'm not sure how you get to the figure of a 10% draw-down, even based on the yearly figures in the above table. Historically Winton's max dd is 25% although of course he has scaled back both his volatility and his profit "targets" in recent years.

rabidric · Post by **rabidric** » Mon Mar 05, 2012 10:23 am

sorry that should have said "something like 10%".

It also wasn't just about Winton, but Large Institutional fund management generally. 10% is just some rough psychological(double digits?) industry concept of "acceptable downside".

Whether is is specifically Max drawdown, or some other form of downside measure isn't relevant. nor is Actual performance from over 5years ago, lol.

The key hurdle is that their projected annual downsides don't appear to the customer to be much over approx 10%.

btw I have nothing against giving the customer what they want, nor anything against the business model of attracting maximum client assets under management and charging them, if that is a viable approach for you.
For most people here it isn't really and so calibrating volatility to a 10% downside floor is impractical, but that is another subject.

Eventhorizon · Post by **Eventhorizon** » Mon Mar 05, 2012 10:57 am

Here's one I prepared earlier ...

Here's an analysis of the change in performance of Winton's Diversified Fund over the last 14 years which I did as an illustration of a Change Detection Algorithm.

Look near the beginning of this post where I set up the problem.

Look at the last section in this post for the analysis.

You don't need to read the whole thing, but you might find it interesting - the math isn't as horrible as it looks!

rabidric · Post by **rabidric** » Mon Mar 05, 2012 11:52 am

I read through those Event. You are wasted on trading, please go solve Nuclear Fusion and Interstellar travel for us first.

marriot · Post by **marriot** » Mon Mar 05, 2012 2:10 pm

Really amazing Ian.
We agree that Winton has cut volatility.
Compare results with Mulvaney.
So i ask again: in your opinion Mr Winton has cut DD or Targets ?
I still suspect it was a kind of " Oh my dear" annual & global results profit targets.
Do you think it is possible?

Eventhorizon · Post by **Eventhorizon** » Mon Mar 05, 2012 8:16 pm

Thanks for your kind comments, Ric and Marriot.

If I had to guess (and that is all I can do) I would say Winton did the following optimization of their trading system:

Minimize volatility subject to achieving a return greater than or equal to the typical pension fund return assumptions.

I think Winton had a revelation similar to that which Ralph Vince talks about in his book "Leverage Space Trading Model" (pp x - xi, p 142) when a colleague, Hiroyuki Narita, states "What is really needed is to be able to maximize not profit, but the probability of being profitable by a certain point in time".

The problem you seek to solve in trading is not "maximize returns" or even "maximize return relative to risk".

PNW_Trader · Post by **PNW_Trader** » Mon Mar 05, 2012 10:18 pm

marriot wrote: i wonder if they have some kind of proifit target whose code is something like "oh my dear, we are up 20 per cent this year, let us exit almost all".
Do you think it is possible code this?

Ralph Vince shows this method in his book - The Leverage Space Trading Model: Reconciling Portfolio Management Strategies and Economic Theory

AFJ Garner · Post by **AFJ Garner** » Tue Mar 06, 2012 4:11 am

Volatility, return and draw down go hand in hand. While it would be nice to be able to increase return while decreasing volatility and draw-down this would not appear to be possible in real life. In real life the trade off seems to be that to achieve lower volatility (and draw-downs that can be tolerated) you have to accept lower returns.

The object of many of us on this forum over the past few years (myself included) has been to increase returns as well as lowering volatility and draw-down. Many of us have assumed that by adding zigs to zags correlation (or rather non-correlation) between instruments and systems will achieve our object. In real life it probably won't. In real life the dream of achieving huge wealth from a small stake by aiming for stratospheric returns is likely to prove elusive.

I suspect it is the same in many businesses. Gearing in property investment can lead to huge wealth in an inflationary environment; it more usually ends in financial disaster when the banks call in their loans during a tough period in the economy. Some people achieve their ambitions - for whatever reason and by whatever means. Most do not.

It is a popular pastime here to deride asset allocators and institutional investors who eschew Mulvaney in favour of Winton. Is it just possible that they may be right to favour lower returns and relative safety of their clients' savings over the dream of shooting the lights out?

SimJimons · Post by **SimJimons** » Tue Mar 06, 2012 6:01 am

Winton publicly stated a couple of years back that they aim for a volatility of 10% going forward. As they have added more strategies (and I would guess even a few short volatility ones) they have become more aware/afraid of "kurtosis risk" and consequently want to reduce average risk. However, they claim that they will be able to deliver the same return, only at a higher Sharpe!!! In addition, which is probably the more important reason why they lowered risk, their clients can't stand the higher volatility they used to run at.

During 2008 they manually lowered risk to protect client money during a period of unprecedented volatility. In addition, they stopped trading forwards and other instruments with a serious counterparty risk. This also reduced risk, or at least that is my guess. Hence, I really don't think that they work with targets on portfolio return level. Instead, circumstances have made them intervene at one point as well as have persuaded them to go for lower risk going forward.

AFJ Garner · Post by **AFJ Garner** » Tue Mar 06, 2012 6:17 am

SimJimons wrote: Instead, circumstances have made them intervene at one point as well as have persuaded them to go for lower risk going forward.

Which, I suppose, is my point really. All that stuff about "old bold traders" and their non-existence is probably right on the nail!

sluggo · Post by **sluggo** » Tue Mar 06, 2012 10:12 am

I guess if somebody really wanted to, they could perform some what-if experiments on past history. They could cobble together a suite of mechanical trading systems, create the little auxiliary blok described below, and then do some backtests on historical data. They could run backtests with the blok "enabled", and then again with the blok "disabled", to see what effects it produced. They could form their own opinion whether these effects were desirable

for a multibillion dollar institutional fund like Winton
for a medium sized ($150M) retail-oriented CTA
for a small CTA just starting out ($2M)
for an individual private citizen, trading her own account

The auxiliary blok might have four user-adjustable parameters

T . . . . the interval of time over which profits are measured
R . . . . the target profit rate
f . . . . the fraction of excess profits that are frobnicated
M . . . . the method of frobnication (a "selector")

Operating as part of the suite of mechanical trading systems, the auxiliary blok looks back over the interval T and measures profits. It might use "Robust Annual Return" (as described in Way Of The Turtle ) to measure profits, or some other statistic. It then computes "excess profits" which are just (MeasuredProfits_over_time_T) minus R, the target profit rate. Finally the auxiliary blok "frobnicates" a fraction of these excess profits; numerically (f x excess_profits) where f is a user selected fraction between zero and one.

What is "frobnication"? It's anything you want it to be. You may want to think of it, very loosely, as "protection". The auxiliary blok "protects" profits. But how does it protect profits? Any way you want it to! When you create the blok, you can invent several different ways to protect profits, and give the analyst a choice of which one of them to test (using the Trading Blox Builder "selector" object). You might want to code and test

Simply liquidate (f x excess_profits)'s worth of existing positions, and withdraw the money from the trading account. Move the money to a completely separate, segregated account and buy risk free, interest bearing assets. Do the appropriate calculations to account for this withdrawal in your VAMI curve.
Use (f x excess_profits) to buy options that protect your existing positions. For longs, buy protective puts. For shorts, buy protective calls. In backtesting you might use historical data of implied volatility to estimate the purchase price of these options.
Move (f x excess_profits) into a countertrend trading system that is negatively correlated to your main suite. If the main suite goes into a drawdown, the countertrend will draw-up and you're protected. If the main suite draws-up, the countertrend will drawdown and equity remains constant. Observe that this particular countertrend system is optimized for protection, not for profitability.
Tighten up the stops on (f x excess_profits)'s worth of existing positions. You will permit them to make additional profits but you forbid losses

I can imagine that you might want to test out T = (time since the start of this calendar year), and maybe T = (rolling 12 month window), or T = (the period of time since the last equity High Water Mark), and of course T = (the period of time since the most recent 20% drawdown).

I also imagine you'd want to horse around with different target profit rates, to see whether there's a "sweet spot" which gives a gain vs. pain profile that you can (A) successfully market to asset allocators; (B) live with; (C) comfortably retire upon

And I'm fairly sure you'd want to invent various frobnication methods M. I mentioned a few of my second-best ideas above; feel free to try out your best ideas.

babelproofreader · Post by **babelproofreader** » Tue Mar 06, 2012 5:30 pm

...It might use "Robust Annual Return" (as described in Way Of The Turtle ) to measure profits...

An even more robust measure would be to replace the linear regression in the RAR% calculation with a repeated median regression (explanation here and here). Some time ago I coded the repeated median (a C++ .oct file) and if anyone is interested I will post the code on my blog and link to it from here.

Eventhorizon · Post by **Eventhorizon** » Tue Mar 06, 2012 7:16 pm

Babelproofreader, you know I would be interested!

rabidric · Post by **rabidric** » Wed Mar 07, 2012 3:26 am

I remember playing around with LAD for a couple of weeks back in 2003 or so, looking at intraday momentum application. It wasn't any good at that application tbh, as it suffers from signal/execution price dislocation just like a DMA approach, for example.

The RMR approach does look like an interesting avenue, especially for applications to performance measuring, though I reckon when it is put to work as a momentum measure it is likely to suffer even more horrific dislocation of recieved execution price from signal point than other regression techniques due to the way the algorithm discards "different" information until it is overwhelming.
I would of course, love to be proved wrong in this instinctive evaluation.

babelproofreader · Post by **babelproofreader** » Wed Mar 07, 2012 8:53 am

Quote:
...It might use "Robust Annual Return" (as described in Way Of The Turtle ) to measure profits...

An even more robust measure would be to replace the linear regression in the RAR% calculation with a repeated median regression (explanation here and here). Some time ago I coded the repeated median (a C++ .oct file) and if anyone is interested I will post the code on my blog and link to it from here.

My blog post is here.

Traders' Roundtable

annual profit target

annual profit target

Re: annual profit target