Thursday, January 29, 2015

An argument against market direction determining your bias

I've been pondering further the question of letting the general market direction determine your trading bias, against focusing solely on the price action of stocks that you are looking to trade.

Regular readers of the blog will know that I made the decision late last year to place less reliance on what the major market averages were doing, as I was not looking to trade them. I look to trade individual stocks based on what their own price action is telling me.

Thinking about this made me delve into understanding how the index prices and movement are calculated, which meant referring to the guide on the FTSE website.(Note: I'm not that much of an anorak that I sat working through all the formulae, I was just looking at the principles of how the various indices are calculated!)

Below are the 10 stocks which have the highest weighting within the UK FTSE100*:

1 HSBC Holdings 7.11%
2 Royal Dutch Shell A 5.53%
3 BP 4.98%
4 Glaxo Smithkline 4.09%
5 British American Tobacco 3.87%
6 Royal Dutch Shell B 3.56%
7 AstraZeneca 3.34%
8 Vodafone 3.23%
9 Diageo 2.68%
10 Lloyds Banking Group 2.47%




TOTAL 40.86%

*from latest figures available from the FTSE, based on closing prices as at 30 September 2014. Currently, there are 101 constituents in the FTSE100.

On that basis, those stocks (basically equating to only 10% of the population of the index) have a combined weighting of over 40%.

Now, as a trader, I want to put the odds and probabilities in my favour as far as possible. The question is, does blindly following the price movements or trends in the index help me achieve this?

As a trend follower, I look solely at the price action of the stocks that are showing the potential for a new trend signal. Each stock I trade is assigned a risk of 1R of my capital. I don't care what their market cap or index weighting is, whether they make widgets, drill for oil, are a bank or anything else - I look at price action. If it is a good set up which meets my criteria, then I will look to trade it, long or short.

Now, consider a situation where those 10 stocks listed above are in uptrends. The other 90% of the index constituents could all be in a downtrend. Depending on the degree of movement up or down in all those stocks and their weighting, it is possible that a small number of market constituents can determine the overall market direction. I'm reminded here of Jim Rogers' comments in his Market Wizards interview about the market action in the run up to the 1987 crash, which alludes to the same principle:

"If you check, you will see that, during 1987, while the S&P and the Dow were going up, the rest of the market was quietly eroding away."

On that basis, if 90% of the market constituents are in downtrends, and yet the market is going up, what would you do? Would you move the odds in your favour if you trade a stock in the 90% rather than the 10%?

Now think about what's been happening in the FTSE100 over the best part of the last two years - no discernible trend. Yet, there have been plenty of constituents have have trended up or down.

The scans that I use will help flag this up for me - the number of potential long or short set ups that come up each day can be a guide. You could even create a scan to simply list which stocks are above or below say their 50 day or 200 day moving averages, and then use that as a quick and dirty guide to determine your bias, which gets around the distortion created by the weighting given to each stock in an index.


To finish, one other history-based question for you to ponder:
  • The Dow Jones Industrial Average was created in May 1896, yet the New York Stock Exchange started in 1792;
  • The headline index in the UK, the FTSE100 started up in 1984. It's forerunner, the FT30 started only in 1935. Yet the modern London Stock Exchange was created in 1801.
If you were a trader and tried to follow trends before the indices existed, would you have been any more or less successful in your trading by not having a major market average as a reference? Would the price action on the stock(s) you were trading have been sufficient? If they abolished the indices tomorrow, what would you do?

Some people may completely disagree with what I've put above, and that's fine. There wouldn't be a market of buyers and sellers if everyone had the same beliefs about the markets...

So, while 'top down' trading has been used very successfully by a large number of market participants, and will no doubt continue to do so, there is an argument which says it is not the only way, and in any case it may not necessarily be the best way to determine your bias.

6 comments:

  1. Excellent post Steve. Can I agree AND disagree? I agree that each stock should be handled on its own merits...once you are in the position already. Where I would disagree would be prior to new entries. I would say that our trading age differs from the old by how correlated stocks are now due to etfs. They say that 4 out of 5 stocks trade with the general market, no doubt there are people out there that could pick that 1 out of 5, but why take the added risk for a new entry with the odds even more skewed against you?

    This is where I would begin to build and hone my watchlist for the strongest stocks, and then when the market clears its "filter" for trend health, I would begin to enter the strongest on my list.

    Could you just use the market as a broad indicator where you would want to be aggressive when the dominant index is in an "uptrend"? I would hold my current positions regardless of market direction once in them, but only take on new risk when the market was in a designated uptrend. This would naturally force the portfolio to be lightening up risk as the conditions get riskier and would be adding risk as conditions improved.

    Am I off base here or does that thinking make sense also? I have debated this issue with myself for years. This so far is the best solution I've come to. As you like say though, if it works for you then it works for you.

    Thanks Steve,
    Cody

    ReplyDelete
  2. Hi Steve.
    I agree. Such a narrow index as the DOW is meaningless. Also, the cap weighting in narrow indexes (FTSE100, CAC40, DAX30, S&P500 yes even S&P) distorts the "average joe" stock which historically and statistically mostly tends to go down on the long run. The indexes being very crude trend following instruments (dump worst loosers, keep winners) have an ingrained survivalship biais, which dont make the task simpler. This study from the CAS Business School (http://www.cass.city.ac.uk/__data/assets/pdf_file/0005/168827/Monkeys-beat-market-cap-indices_Final030413.pdf) clearly demonstrate that the arbitraging resulting from overweighing narrow indexes in a majority of portfolio makes market cap weighted indices perform worse than any other weighing method (including throwing darts). So definitely, a market depth measurement has to deal with very large indices, and if possible not do market cap.

    As trend followers, we need (imo) pay attention to the global trend relative to our timeframe (weekly if I trade daily charts, etc...) to not overweight a portfolio counter-trend. To be honest, I do not short stocks in a global uptrend and that's it, I only take hedges on a global index. When the trend reverses, then its time for me to go short, even if shorting stocks is not my favorite sport. Also, I scale my position sizing on a very short moving average of my equity in terms of R lost/gained (so that if 4 trades in a row are winners and i come out of a string of loosers i dont wait 2 months to load back up).

    I'm not balsy enough to not apply a trend filter atm :).

    Thanks for all the good posts.

    ReplyDelete
  3. Agreed. Good post.

    What we really have is a ‘market of individual stocks', not a stock market. At any given time you can find trending stocks, regardless of where the broader market averages are.

    You may also notice the longer term breakouts tend to lead the averages. Many shorter term traders seem to use the broader market action to filter out some of the noise. But I think that's a complicated way to trade. The longer term trend discounts all one needs to know and for me that trust came with time and experience.

    ReplyDelete
  4. I agree with Steve. This is a fat tail system with a skew on the right i.e. 95% of profits come from 5% of the trades. Those 5% trades will tend to shoot up and do what they want, and act strong as they have their own story. Therefore if you dont take a strong acting signal as the overall index is lagging, you might miss all your years profits. (cherry picked) example would be 2008 - family dollar store vs the S&P500. a lagging index is likely to cause more breakout failures and for example a weak breakout in a up market might lead to 0.5R gain where as 0.5 loss in weak market. whilst they would add up and impact performance/win rate, the point is that 5%/95% effect.

    ReplyDelete
  5. I think the strongest argument in favour of ignoring the indices is what you say about shares trending over the last few months despite no clear trend in the FTSE. And if anyone knows how to identify these trading opportunities, please let me know. But in my experience, a share may well break out from the 20 day high in unison with the index but if the index reverses, so does the share price. Not sure if this is because I trade predominantly smaller cap shares which may not carry as much momentum when they break out?

    ReplyDelete
  6. Thanks for all the comments guys. Obviously struck a chord with a few people as I normally don't get this many comments on a single post!

    ReplyDelete