Money Management

12th May '16 by roddy10

16 comments 6780 reads

Let me start with a question - which of the following will perform best:

1. A careful selected portfolio of 20 stocks?

2. A portfolio of 20 stocks that are chosen at random but with strict money management rules?

I am a value orientated investor but I would suggest that the performance of portfolio 2 might surprise many people. A lot of modern investment management is focussed on the process of stock analysis and very little is on money management.

So what do I mean by 'money management'?

Money management is a set of rules to systematically size positions, control risk and put on positions and take them off in a structured, systematic and planned manner.

Often the best way to start a discussion is by discussing what one does - so here goes.

1. Stock selection

The mechanics of how to select a stock are varied but generally fall into 'story lovers', ' technical analysis', 'fundamental analysis' etc. The key point is recognising what type of investment it is and WRITING DOWN your thesis.

2. Work out how much you are prepared to risk

I am generally happy to risk 10 - 250 bps per position. Others will have different mileage. For me what matters is (i) how much work I have done (ii) how much 'edge' I have (iii) how does the risk compare with the reward.

So I do not look at risk in isolation but also in comparison to the reward. (Aside: In Van Tharp speak I look at the expectancy - except I look at the probability weighted expectancy).

Again I write down my estimates.

3. Work out a reasonable place to have a stop

There are lots of different methods people use to place stops. I tend to use a 'soft stop' and a 'hard stop'. The soft stop is placed at a level where I will question my hypothesis. It might be a couple of standard deviations from the current price or it might be at a clear level (eg for a company that has recently IPOed I know other investors will get nervous if it falls back down to the IPO price) or it might be based on a seasonal or technical issue.

I am sure the fundamental investors on this site will 'poo-poo' the idea of a stop loss. I generally spend a long time researching stocks - but I am also grey haired enough to know that sometimes I get things wrong. Indeed one of…

Unlock with Email

Unlock with Google

Already have an account?
Login here

roddy10

12th May '16

(Heck I hope I have not scared everyone - would love comments - even if to say it does not make sense!)

Ramridge

13th May '16

Hi roddy -
Good write-up. May i suggest that a better title might be ' risk based investment strategy' rather than 'money management'?

Now, oddly enough I have just finished reading 'Momentum Masters: A Roundtable Interview with Super Traders' by Mark Minervini and other very successful traders. A lot of what you write is similar to the methods used by these luminaries.
The book is a bit pricey but well worth it.

Mancman

I found this very useful. Applying a set of rules like this would have saved me a great deal of capital loss in the past two years!

fizzlebiz

Yes, "money management" (as you call it) is a skill in itself. I am aware of managers that have relatively poor alpha generation but have outperformed due to "money management". However, your system relies on the ability of the individual investor to quantify risk. I would bet, therefore, that most investors would find it easier to manage the risk of a portfolio of stocks they know against a randomly selected one. Either way, the value of this system seems, to me, unverifiable (although it is similar to the one I use myself).

I also don't think the portfolio you propose is well-diversified at all. The aim seems to be to propose a set of companies that are all individually different from each other, this results in a diverse portfolio but not diversification. For example, a portfolio consisting of Merck and Telecom Argentina is diverse but not diversified. Diversification, conventionally defined, means owning a little of everything and earning the undiversifiable risk premium associated with each asset class. I get the point you are making, correlations change and stocks that seem dissimilar can often by exposed to the same factors (measuring factor exposure is, after all, what risk management is all about), but minimising sector risk isn't the same thing as diversification.

Mark Simpson

Generally a lot is written about how to pick great stocks and very little on how to build those into a portfolio that makes the best out of stock-picking skill. The latter always somehow seems more boring than the former despite it's equal importance in long-term returns.

This is how I try and approach portfolio management:

Where risk is usually a combination of the following sort of things:

Financing risk – Companies with high levels of debt will be less able to weather a period of poor trading. A heavily loss making company may struggle to raise extra capital.

Management risk – Management with a history of poor capital allocation may not act in the best interests of shareholders.

Product risk – Companies with one product or a few products serving only one market are more exposed to the performance of that product or market. Typically, but not exclusively, this makes smaller companies riskier.

Liquidity risk – The risk of not being able to sell near the published price when you want to. Typically, but not exclusively, this makes smaller companies riskier.

Commodity risk – Companies without a sustainable competitive advantage to control pricing are more likely to be exposed to commodity, economic growth, inflation or obsolescence factors.

Correlation Risk – I don’t mean the mathematical co-variance of stocks but the common exposure they have to factors that are out of your control. For example if you already have an oil explorer in your portfolio adding a second one should be considered higher risk (all other things being equal) than adding an equally undervalued oil consumer such as a plastics manufacturer.

More details here: http://www.dangercapital.co.uk/2015/08/TheMatrixRe...

Obviously the assessments themselves are subjective but I find having a disciplined framework like this to work with stops me making major portfolio construction mistakes like over-sizing high risk punts or under-sizing when a truly 'fat pitch' comes along.

1. A lot of my ideas on money management come from a combination of four sources - (i) casino games - particularly roulette which ought to be a pure game of chance (assuming you are not Doyne Farmer or a relative of Mick Jagger) (ii) Van Tharp (iii) experience (iv) books

2. I have thought about running a computer simulation using money management techniques and a randomly selected portfolio but never got around to it. (Though I did once use all the components of an index but it was a long time ago). I also used money management techniques to model playing the evens chances at roulette in my youth.

3. I think your point re diverse but not diversified is sublime (ie very good and subtle). It has really made me think - you are absolutely correct. But I would suggest that many people confuse the two - which indeed is the point I was trying to make. As you highlight a diverse set of sectors is not the same as diversification - and indeed that is the issue. The problem is that without knowing the stocks and the underlying businesses well it is usually a post-hoc analysis that shows they was not diversification.

Danger:
That matrix and your website is extremely impressive and I shall endeavour to incorporate it into my process. I am truly humbled by it.

I would add one more thing to being disciplined - which is perhaps aimed more at new investors:
(1) The biggest secret in investing is TIME
Much of most days I do nothing but read. Most novice investors feel they need to trade every trade - you don't - and you don't need to screen watch every second (unless u are a highly levered day trader). Secondly if you are too busy spatting flies you will miss the fat canary. Thirdly it takes time for a company to generate returns.

DWit199

14th May '16

The whole point of Stockopedia's ranking tools is to identify shares that are likely to generate a higher than average return at lower than average risk. If that is your investing universe, then equal position sizing is a logical way to construct a portfolio. Much academic research is based around a 1 year rebalance which is a reasonable default position.
I think you are complicating things too much.

16th May '16

roddy,

Thanks. I fully agree. Reading widely is one of the keys to success. Have you come across this:

http://fivegoodquestions.co/

It is a guy who was inspired by a chance lunch with Warren Buffett to read more and now interviews authors of the books he's read to ask them 'five good questions.' It's a great source of new book ideas.

David,

I also fully agree that equal weight is the way to go if you are unable to differentiate between the risk or return profile of stocks. This is clearly the case when using just Stockranks for your portfolio construction. However by doing this it is clear that:

a) You are making portfolio weighting decisions because you are setting a limit where your allocation goes to zero. You hold a stock with a stockrank of 90 and not with 89 for example, though in reality there may be little between them. Or by choosing the highest ranked share in each sector you are weighting within sector.

b) If you can consistently make return or risk judgements that are mostly correct you add significant value by weighting your portfolio even within a pure quant strategy.

Now it is perfectly valid to say that you are not able to do this. However at it's most basic application if you can (using qualitative or quantitative data) exclude the likes of Globo or China frauds then you have added significant value. So it may be worth trying.

I reckon most experienced investors are capable of making a judgement about the risk profile of a given business. What leads us astray is not the inability to correctly categorise the risk level but our willingness to ignore it when the perceived upside is large!

There is empirical evidence to support "using just StockRanks" with an equally weighted portfolio rebalanced at regular intervals. There is very little, if any, to support adding a further quantitative or qualitative judgment to those StockRanks.
It is uneconomical to invest in all shares with a StockRank >90 (or whatever limit you choose) so my "weighting" process is really an attempt to generate a representative sample at reasonable cost.

wattanut

17th May '16

Hi Roddy

I pretty much agree that management is at least as important as stock picking. I'm not sure I would go as far as choosing completely random stocks though.

Risk / reward is interesting, does anyone consider the overall state of the market when calculating it? Before I think about buying anything I look at the 50 and 20 day moving averages on a one year and three year FTSE All Share ASX+ chart. Other than a few positions in gold and defensives, I have been out of the market since last June when the index price dropped through the 50 day ma. Since then the 50 day ma has passed through the 200dma and has yet to recover. On the three year chart we have just had a 50 / 200 death cross.

Even the 90-100 Stockrank performance chart for companies over 50M is at the same level it was a year ago. Given that 25-65% of share performance is down to market conditions (depending on which book you read), where is the reward for the risk to your capital?

You may have guessed, I hate losing money!

There is empirical evidence to support "using just StockRanks" with an equally weighted portfolio rebalanced at regular intervals. There is very little, if any, to support adding a further quantitative or qualitative judgment to those StockRanks.

Logic says that if you are able to weight even slightly by upside potential or downside risk then it is worth doing. A lack of empirical evidence on the benefit on a particular quant methodology we call 'StockRanks' doesn't mean that it doesn't improve returns. There are other empirical studies on other quant methodologies that show that it does improve returns. The early work of Bob Haugen in the field or Gray & Carlisle removing potential earnings manipulators in Quantitative Value are just two that come to mind.

It is uneconomical to invest in all shares with a StockRank >90 (or whatever limit you choose) so my "weighting" process is really an attempt to generate a representative sample at reasonable cost.

Unless you are doing this sampling with a random number generator then you are putting your subjective decision making into the mix. Having simple rules like avoiding foreign stocks on AIM or trying to exclude companies with the cash flow features of Globo or Quindell would add value.

Don't get me wrong, I think equal-weight is a good idea when following a quant strategy (like StockRanks) that intentionally says 'I am unable to make any further valid judgement about the investment merits of this company.' Just most investors are not like that and to suggest equal-weight is the ideal for all investors who are capable (or at least believe themselves to be) of making these judgments is just not true. As you point out even investors who accept that they don't know and would like not to have to make any judgement on individual companies end up having to do so for economic reasons. If you are going to have to do it anyway you might as well have a strong logical framework to do so.

A lack of empirical evidence on the benefit on a particular quant methodology we call 'StockRanks' doesn't mean that it doesn't improve returns.

The converse is also true.

The early work of Bob Haugen in the field or Gray & Carlisle removing potential earnings manipulators in Quantitative Value are just two that come to mind

The QV book was produced to support a commercial product, so hardly unbiased. It may work but the proof of the pudding will be the performance of their ETFs. I think it is too early to form a judgement. The same objection could be leveled at Stockopedia but at least the prospective component of the project is being performed openly and transparently. It is worth remembering that theStockopedia Quality Rank includes Beneish and Altman scores so the issue of earnings manipulation and bankruptcy risk is addressed to some extent.

Unless you are doing this sampling with a random number generator then you are putting your subjective decision making into the mix.

A minimally biased representative sample is not necessarily totally random. This particularly applies where the population under study has easily defined subgroups with different properties that can be sampled in proportion.

The converse is also true.

No it isn't. Logically if you can overweight stocks with a higher expected return your portfolio will outperform equal weight. You don't need empirical evidence, simple logic will suffice.

What you are doubting is that it is possible to do so consistently. Now that is valid critique. My argument to that critique is that is has been done successfully on other quant methodologies so may be possible with StockRanks but I accept it is far from proven. The benefits outweigh the risk IMO particularly when it comes to avoiding frauds but accept that not everyone will hold the same view.

It is worth remembering that theStockopedia Quality Rank includes Beneish and Altman scores so the issue of earnings manipulation and bankruptcy risk is addressed to some extent.

I think this is one of the weaknesses of the StockRank implementation. When you have earnings manipulators they tend to have very high quality scores since the other factors that go into the QualityRank - GPA, ROCE etc. - tend to be much higher than similar companies. If you are going to make the numbers up then you might as well make them look great right? Much higher margins or ROCE without a clearly identifiable sustainable competitive is actually a red flag not a positive for the company. If a company is likely to be manipulating it's numbers then it makes the whole QualityRank unreliable. That's why including Beneish in the QualityRank isn't that effective in this regard. It is much better as a screening tool than a ranking tool.

18th May '16

David:

This is something I would love to have the raw data to test.

It seems to me that there are a number of different issues:
1. Stockopedia ranks provide a means to find stocks which, are _believed_, to generate higher returns than the general investing universe. The reason I use _believed_ is not a criticism of Stockopedia but to show that _generally_ quant models and their performance is _usually_ based on empirical observations and in some ways retrospective data-mining e.g. a quant analyst will develop a model and see if it has worked on historical data. (Ok I know that Stockopedia has run realtime prospective models but I am trying to make a general point).

2. What we (or at least I) don't know is whether the performance by sector is equal. Additionally even the prospective models show differing performance by region. To give an extreme example if a tech company in the US and a utility in Germany had equivalent ranks would that lead to equivalent risk (i.e. either volatility or risk of drawdown) and equivalent return.

3. Ideally rather than equal weighting, I would suggest one would want to allocate an equal risk budget by position. Going further I would suggest that one really would want to allocate to positions such that the reward vs risk was equivalent. However then one would want to adjust to ensure that there were not embedded correlations (e.g. too much Chinese exposure).

I was wondering about a point that you raise. How would Enron, Worldcom etc rate on a ranking system. Or actually would they have ranked highly on the way up but fallen out just before they turned?

Similarly I have often wondered how well quantitative systems (including Stockopedia) would have done in the dot.com boom? You will recall the story of Tony Dye - would a quant ranking system have also 'lost its job' for underperforming ?

Money Management

Unlock the rest of this article with a 14 day trial

About the Author

Roddy10

16 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...