Low-Risk investment strategy ?

Aegis

aliciathyme wrote: »

All the results I have posted have been the result of rigorous modelling of 5-year periods from a FTSE 100 constituents dataset of daily prices over the last 11 years. The data covers periods of rising, falling and static markets, and I have developed two specific methods to select potential buy targets - the Rising Trend process for selecting an equity for which the share price shows a sustained rising trend, and an Anomaly process which looks for exceptionally volatile shares and works best when the price of an equity is falling.

However, the processes are not geared to the general market rise or fall as such as they work on individual equity prices and these may or may not follow the market.

So, the results of my model for the last 5 years are not based on an arbitrary set of parameters, but parameters which have been carefully selected to produce the best results over the last 10 years.

The trouble is that it sounds arbitrary, in that you have made changes which would have achieved better results in the past without explaining your justification as to why these changes should have positive effects going forward. In essence, you are assuming that past performance is a reliable predictor of future performance based solely on technical analysis of stocks.

I could probably design a system in an afternoon which would generate higher returns than yours in back testing because I know what the outcome is before I begin. All I'd need to do would be track one volatile stock and tweak the buy/sell parameters until huge returns were modelled. I wouldn't ever try to implement such a strategy going forward though, as the parameters would have been entirely arbitrary and optimised solely for past performance.

These are the results of the model options for the current 5 year period to 15 May 2013:

Original Scheme: 40% Stop Loss (no take profit or early sale):
Rising Trend process: Net Profit = 116%
Anomaly Process: Net Profit = 32%
Both Processors: Net Profit = 132%

Revised Scheme: 33% Stop Loss, early sale on Stop Loss 10% within first 2 weeks, Take Profit at 10% fall after at least 33% profit
Rising Trend process: Net Profit = 233%
Anomaly Process: Net Loss = 3.4%
Both Processors: Net Profit = 258%

The point of modelling is to permit continual development of the processing options as new data becomes available, always back testing over a long period (currently 11 years) to confirm any improvements.

Modelling has its uses, but you shouldn't fall into the trap of believing that optimising past data is the way to generate the best ongoing returns. The books I've read on designing an investment system all warn about the dangers of over-optimising against past data, as the tweaks may only work on that past data rather than on future data.

You have to devise trading rules with solid justifications, regardless of whether your system is based on technical analysis or fundamental analysis. Simply optimising past data isn't enough.

As you can see here, operating solely with the Rising trend processor (which I have previously explained in detail) gives exceptional results for the current period (the best of 4400 commercial funds is currently Fidelity UK Smaller Companies at 188% return). Of course there are no guarantees for future performance, but the results have been consistently good over the last 10 years..........................

Again, this means nothing because you have optimised your system to perform well using the known data over the past 10 years.

I'm not trying to put you off developing your own system if you want to, but please bear in mind that you have had the benefit of knowing exactly what will happen when designing your system and running the past data through the selection process. You have had the benefit of tweaking the system based on large numbers of experiments on the past data. Going forward, you only ever get one shot at implementation on new data, and you will already have seen the wild differences you can achieve with only minor adjustments to the decision parameters.

If you're confident, then run your own money according to your model. But please don't keep trying to promote this as a low-risk strategy; that's simply not true. This is a risky trading strategy based on a technical analysis of past data which is subject to revision periodically to try and chase enhanced returns.

aliciathyme

Aegis wrote: »

The trouble is that it sounds arbitrary, in that you have made changes which would have achieved better results in the past without explaining your justification as to why these changes should have positive effects going forward. In essence, you are assuming that past performance is a reliable predictor of future performance based solely on technical analysis of stocks.

I could probably design a system in an afternoon which would generate higher returns than yours in back testing because I know what the outcome is before I begin. All I'd need to do would be track one volatile stock and tweak the buy/sell parameters until huge returns were modelled. I wouldn't ever try to implement such a strategy going forward though, as the parameters would have been entirely arbitrary and optimised solely for past performance.

The changes I suggested are the result of some peoples' anxiety over being trapped in the 40% Stop Loss regime and missing out on new buying opportunities, and also the problem of missing out on a good return where a share price dropped from a peak before a new Target Buy initiated a sale.

Nothing has changed in the basic strategy of Sell-to-Buy and the target buying selection processes have remained the same. It is only the Stop Loss policy and the Take Profit function which have been revised.

In both cases tested, there are the same number of Target Buys (158) but for the 40% stop-loss scheme only 56 could be taken up due to holding more losing positions compared with 82 taken up for the newer scheme. The win ratio was lower (67%) with the newer scheme compared with 82% in the 40% version. These effects are exactly what one would expect from these changes.

Incidentally, I do not involve forward prices in any of the strategies, and whilst your comment "I could probably design a system in an afternoon which would generate higher returns than yours in back testing because I know what the outcome is before I begin" may well be true for you, it has no relevance to my modelling.

Alicia

aliciathyme

Today sold Hargreaves Lansdown @ 18.2% net profit to buy POLY

Alicia

doughnutmachine

aliciathyme wrote: »

Today sold Hargreaves Lansdown @ 18.2% net profit to buy POLY

Alicia

if this was dragons' den someone would be asking your background.

perhaps it would be easier to "sell" this idea if people knew more about you? for instance do you have a degree in maths or something? or a career in fund management?

i have to admit being sceptical to your method.

Jegersmart

aliciathyme wrote: »

Today sold Hargreaves Lansdown @ 18.2% net profit to buy POLY

Alicia

POLY.L (Polymetal Int)?

aliciathyme

Hi J,

Yes, Polymetal International, it's the only POLY in the FTSE 100!
(it came up as a but on my anomaly processor)

Alicia

Jegersmart

Right, so not a trend trade, an anomaly trade. I guess there have been other buy signals recently? ARM, LLOY and SL perhaps?

J

gadgetmind

doughnutmachine wrote: »

do you have a degree in maths or something?

What, like all those guys at Lehman Brothers did?

JohnRo

gadgetmind wrote: »

What, like all those guys at Lehman Brothers did?

Tangent...

Have you watched the beeb2 documentary called bankers?

I thought the first one was a bit superficial and light touch but the second was much more interesting.

The general theme of risk and how all bought into the idea that with a big enough spread sheet and enough calculations they could see into the future, I wonder how much has really changed.

Jegersmart

gadgetmind wrote: »

What, like all those guys at Lehman Brothers did?

A much better example of this is LTCM....

imho

J