Efficient Market Hypothesis

In reply to AbAngus (post #22)

Presumably, all HFT programmes have pretty similar algorithims - so there must be a risk of all HFT programmes wanting to the same thing at once and thus creating sudden large swings....?  Are we about to enter an age of super volatility?

Yes I think that is precisely what Myners is (rightly) warning against! I am reminded of the impact of portfolio insurance being coupled with the market discontinuity (ie the hurricane) in 1987, which I am entirely certain was responsible for the extent of the fall on that occasion (because the hedging algorithms created a one-way market as the models kept trying to rebalance their hedges).

I noted an example of this earlier today elsewhere, in relation to trading in Dana first thing. I copy my comments below:

There was an absolutely astonishing piece of trading by someone just after the open today. Have a look at the trades timed at 8:02:25 here:
....with the best bid then at 1267p, someone blasted out over 90,000 shares in an instant and hit every bid on the book all the way down to 1150p!!!!
I can only think this was a cock-up!
Nobody surely can have intended to hit bids that were over £1 lower than their starting point! ??
....an unexpected windfall though for those with bids in at between 1150 and 1220 or so.... ;-)

ee

An interesting article from the BBC, encapsulating a warning from Paul Myners:

http://news.bbc.co.uk/1/hi/business/8338045.stm

Companies could become the "playthings" of speculators because of super-fast automatic share trading, Treasury minister Lord Myners has warned.

HFT [High Frequency Trading]  is based on computer programmes which, in effect, remove the human element from share transactions by searching for trends in the market and trading automatically in a fraction of a second.  It is estimated that HFT has risen sharply in the USA and now accounts for up to 70% of all share trades and is rising rapidly in the UK and Europe.

Presumably, all HFT programmes have pretty similar algorithims - so there must be a risk of all HFT programmes wanting to the same thing at once and thus creating sudden large swings....?  Are we about to enter an age of super volatility?

AA

In reply to SW10Chap (post #20)

It is a good point but it is very difficult to prove it across all the markets, if by all markets you meant Real estate, commodities, derivatives etc. In principle, these arguments should be applied across all the markets. But the studies I am referring to are basically around stock with empirical evidence from stock market.

Recently there are studies about AI/algorithmic trading also, I have been through some of these. I am not convinced about AI etc. Also, lest we forget, what happened with LTCM which had a high profile team of PHDs and another good example is of Renaissance Capital which has an army of computers (one of highest IT spend) and does a lot of algo trading is now looking for capital injection.

RH Thaler says that markets are not efficient; opportunities get created due to anomalies (behavior). But Mark R questions can these anomalies be exploited, there is a cost to arbitrage (e.g. what you said about viscosity in small cap stocks) and he proves one can get very good returns just by doing passive investing over a long run. Please remember active investing is very expensive and rarely all the nice data reflects the true costs of it & most of the studies suffer from the survivorship bias.

In reply to puneetk (post #19)

I like Mark Rubenstein argument on why Markets are nearly efficient.

Outline it for us then. Is that for all markets, or just a subset?

I have to say I'm far from convinced and it's a shame tha the Invisible Man did not respond to to my query as to whether he thinks AI will be able to read between the lines of a technical RNS and decide to what extent the news is god or bad (post #16). I'm genuinely interested to know whether people believe that AI will provide these kind of answers in future - because the market certainly doesn't appear efficient today.

SW10

Yes I am taliking about the same R Thaler but, in haste, I have mentioned incorrectly - it is not R Thaler but Mark Rubinstein. My apologies.

R Thaler strongly believes that markets are inefficient and is a strong advocate of behavioral finance.  There are few debates on this by  Thaler & Fama, Thaler & M Rubenstien (you can google it). Lot of people have analyses those (http://www.duke.edu/~alexrose/AlonREMpaper.pdf)

I like Mark Rubenstein argument on why Markets are nearly efficient.

In reply to puneetk (post #17)

You mean this guy? http://en.wikipedia.org/wiki/Richard_Thaler - have you got a reference for the research?

If possible try to read artciles from RH Thaler. He puts is his case very nicely on Markets being nearly efficient and explains with examples that Indexes (I things he refers to S&P, Vanguard an few others) have out performed the fund managers over a period of time.

In reply to The Invisible Man (post #15)

Don't count out EMH yet.  Won't be long until the efficiencies at edges - in small caps and value stocks - are ironed out too.

Whlst I recognize that you probably closer to it than I am, I really don't see that happening in my lifetime in my area of interest of small oil E&Ps. For example, I cannot conceive of an AI that will be able to read between the lines of a technical RNS and decide to what extent the news is god or bad. Just as a quick illustration: http://www.investegate.co.uk/Article.aspx?id=200909211452364022Z

Buy or sell?

I don't believe you'd get a uniform view on that - or many others. And if there isn't a 'correct' answer that can easily be drawn, then howw will the AI ever be able to learn?

My sense (self-deluded or not!) is that the few people who can read between the lines of those kinds of announcement will always be a few days or weeks ahead of the herd.

SW10

Many of our technologies are extremely crude at present. But technological change is increasing in pace rather than decreasing.  Artificial intelligence is only a few decades from matching human intelligence, after which it's powers of parallel processing will far exceed the limitations of biological computation.  I already work with crude AI, and it's already more market intelligent than most human traders I've come across - more responsive, more receptive, more profitable.

Don't count out EMH yet.  Won't be long until the efficiencies at edges - in small caps and value stocks - are ironed out too.

I can understand that computer driven trades could be fairly efficient - but human judgement depends on the brain.  The brain is mostly water - a substance not normally known for its intelligence.  True, there are bits of fat and protein - which presumably add a bit of flavour but not much else.   It is amazing to me that the brain can get so many things 'right' - but it is by no means 100% efficient. 

LFC makes the point that lots of people being 'inefficient' in different directions may still be efficient as a whole - but history suggests that humans are prone to lemming like panics or enthusiasms....  

While humans remain in chrge the market will not be entirely efficient (thank goodness).

There was an article about the EMH is yesterday's FT: http://www.ft.com/cms/s/0/ec77ee24-b2aa-11de-b7d2-00144feab49a.html

Markets after the age of efficiency

By John Kay

Published: October 6 2009 22:00 | Last updated: October 6 2009 22:00

Warren Buffett said most of what you need to know about efficient markets. “Observing correctly that the market was frequently efficient, they [academics, investment professionals and corporate managers] went on to conclude incorrectly that it was always efficient. The difference between the propositions is night and day.”

The difference between these propositions is also the difference between a $50bn fortune and the returns of the average investor. Mr Buffett has made his money not from the part that is frequently efficient, but from the part that is infrequently inefficient...

 

On the subject of Investment Trusts... (marben) - I couldn't agree more.

I held stock for many years in Capital Gearing Trust PLC, which showed about 25% compound annual returns over a period of nearly 30 years.  It's still a relatively small fund, but has grown from almost nothing - anyone with a calculator can do the maths.

It was run by Cazenove chap for almost a generation, he leveraged when long term valuations were low, and came almost completely out of stocks every time things got  too frothy. There were periods of underperformance - e.g. in 2005-2008, but it never had a calamitous drawdown like most funds.

Now that's real returns, and an understanding of risk  which noone in the city seems to recognise any more. Risk is about downside protection - Valuation risk, business or earnings risk, and balance sheet risk.  It is not about beta, or volatility.... the young quants have a lot to learn - and the market will dish out lessons in spades the longer they hang around.

PS - I am no longer in any way connected with CGT PLC ... but for those interested... http://www.capitalgearingtrust.com/

For EMH to hold we have to assume that the actions of millions of individual traders/investors sum up to efficiency even though we all know that the individuals themselves are far from efficient. And that the market operates as a cooly logical information-processing machine rather being driven by the all too human emotions of greed and fear. We also have to explain how manias develop into bubbles give way to crashes etc.

Nope, EMH doesn't hold much water in my book.

lfc

Good responses chaps.

I particularly like SW10's analogy.  Of course I'm not even going to try and discuss it with my tutor because :

a)  They aren't likely to know what I'm talking about

b)  I'll sound like an arrogant tw*t (are we allowed that word on here? ;-)

Surely thats where the  Oily technical community can have a lead on the analysts, we can take our experience and ability to make judgements on predrill COS, likley reserves and flow potential and take a risked position in a company thats drilling a well that could be "gamechanging" to the share price.(assuming they put enough information out)

Short of leaked information/rumours (like GKP yesterday)  we can at least sit  more comfortably with stock predrill.

Thats probabally why the techies are frustrated with certain stocks currently a bit slow with the drill bit as it will take a well, or other tangible piece of technical evidence to move the  share price(up or down).

Flyinghorse

Someone once illustrated EMH to me by saying: "there can't be £10 note on the floor over there, otherwise someone would have picked it up."

I have some sympathy with The Invisible Man's view that The market is always right - after all, it sets the price. This, to me, is one of the great Catch-22s of investing, particularly with small-caps. If a company has a hidden treasure-hoard which I find out about through my research and no-one else does, the market will never reflect that price. Of course, that's not an efficient market, but as (I believe) most here expect to profit from inefficiencies our ultimate hope is that that knowledge will emerge and the market will eventually have 'perfect' knowledge.

Having spent a significant part of my working life with fluids of different viscosities, that's the way I tend to think of it. Knowledge about most big-caps tends to have low viscosities (highly fluid) and so corrections to price are made quickly. Knowledge about small-caps tends to be more viscous, and so there is opportunity for the informed, the knowledgeable (and perhaps the more brave?) to take advantage. Unfortunately, I've also learned that there are cases where knowledge is so viscous that those corrections take a long time to arrive!

So, I don't believe in EMH for anything but the biggest caps where knowledge is highly-fluid (although I take Marben's point about the banks) and prefer to look for stocks which I believe will have a degree of viscosity - but not too much.

Say, an API gravity closer to 30 than to 10...

SW10

[Edit: unintended double-negative removed]

Most people aren't.  I'm looking at this from a short term trader's perspective. All the best short term traders are on city dealing desks using quantitative methods

Completely agree: amateurs have little chance of competing with the pros if they try to trade short term. With techniques like HFT; the commission structures institutional traders enjoy; and immediate access to all the latest news & analysis (v few non-institutional traders pay the hefty fees needed to get that), it isn't exactly a level playing field. However, when it comes to INVESTING, that's a slightly different story. ;0)

Incidentally, not even large-caps are particularly "efficient" - look at the SPs of the major UK banks since the start of the year, for example. Have things changed so much that a quadrupling of Barclays market cap is justified? More a case of technical factors (& sentiment) causing distortions. Such distortions, positive or negative, are constantly present, to a greater or lesser degree.

Cheers,

Mark

The real paradox to me seems to be that the (relative) efficiency of the market relies on people believing that the market is inefficent, doesn't it? Otherwise, what's the point of putting the work in? In any case, I think these EMH studies are distorted by the impact of the large-caps, which are largely efficient IMO. There's plenty of fertile ground to hunt for inefficiencies and mispricings amongst the small caps. That's plain as day, surely?

Ok - the defining factor there is

 I am prepared to put considerable effort into research & study.

Most people aren't.  I'm looking at this from a short term trader's perspective. All the best short term traders are on city dealing desks using quantitative methods. Every private investor out there using 'breakouts' or qualitative measures such as Elliot Wave to time their trades are toast.

I would suggest that there are long term structural inefficiencies, which lead to trend trading and value investing being profitable, and may be the reason for your outperformance.  If you can discern those structural or value inefficiencies then you may be able to garner an edge - but even that may be randomness.  Your 12 year time frame is probably not long enough to prove you have an edge - what has the voltatility of return been?

Better to take one's ego out of it, and accept the market is largely efficient.  Stomping one's feet for a year when the market gets it wrong and you are out of the money doesn't change the fact that the market doesn't care.  The market is always right - after all, it sets the price.

Price is everything. Value is nothing. Has been since Oscar Wilde's days.

If anyone needs evidence that the EMH is bunk, I'd suggest studying the market for ABCP. All the information was there, for anyone that took the time to study and analyse it, yet that paper was evidently mispriced for a long, long, time. The ABCP market was not exactly small and it was, supposedly, a market for professionals.

InvisibleMan: what returns has a tracker made over the last 12 years? I have made an IRR on my self invested ISA over that period of 11.6%p.a. My money has trebled. My SIPP, which is more broadly diversified, has made 13.6% p.a. over the last 7 years. I most certainly don't think I am an exceptional investor (though I'm working on it :0)) but I am prepared to put considerable effort into research & study.

For a period, I included a tracker in my SIPP but, as I found that I was consistently beating it, I ditched that element a couple of years ago.

For those without the time for detailed research into individual companies or investment instruments, I'd suggest studying some well run investment trusts, such as RCP, which has also outperformed the market over a long timeframe. [I hold that one in my SIPP]

Cheers,

Mark

The market is generally efficient.  At least efficient enough to ensure that fund managers in aggregate cannot outperform simple tracker funds.  But it's not efficient enough to stop arbitrageurs from taking their edge.

I work in the hedge fund industry and we make a lot of money by providing liquidity to fund managers when they have to find it.  If we are making money, then they are losing it.  Sure we are tiny in comparison, but every incremental edge that is lost by the elephant funds eventually brings them to their knees.  There's a world of parasites out there, who are taking the small inefficiencies and running with them.  They are on every dealing desk in the city, and they are taking funds for a ride.

For small investors, of which I presume there are a lot on this site -  save your time, and your money, and just buy a tracker - you cannot compete.

The city loves EMH , and the CAPM (Capital Asset Pricing Model) because it allows them to quantify 'risk'.  As soon as you quantify something, you can hire dozens of expensive overeducated individuals with streams of letters after their names. 

Together they build departments with imposing sounding names like the 'Market Risk Analysis Professionals' and  throw around wonderfully obscure acronyms like Value at Risk (VaR), Stress Tests and so on.  New Professional Qualifications are built around teaching these acronyms, and suddenly CEO's are surrounded by youngsters sporting snappy suits and even more impressive sounding letters after their names.  They all get so impressed with themselves that they actually start believing this bull####.

These industries have been based upon the idea that you can quantify the risk of something by measuring its standard deviation... who the hell ever gave a damn about upside volatility?  What moron first came up with the idea of basing an ENTIRE INDUSTRY'S risk management process on how much something jiggles around.

It's all nonsense and classic groupthink. Each evening while I tuck my kids in with a rendition of 'The Emperor's new Clothes' I chuckle very quietly.

Hello Nigel,

I was taught the EMH 25 years ago at business school. I then started working in the City - and very quickly realised that it was utter bollocks.......

......but that is only true when looking at specific stocks from the perspective of someone in the City - and it is the City's JOB to make the markets as efficient as possible for those who are slightly less fleet of foot! Over the market as a whole, they don't do a bad job of this, on any timescales beyond a few days.  Therefore whilst one can certainly see successive opportunities to earn excess returns by making trades on the back of newsflow if one is paying very close attention and standing ready to deal it is also true IMO that Joe Public (reliant on newspaper tips etc) would very much struggle to beat the market.

I suspect that most people here would think they were a bit sharper than the market, on average...which is probably as it should be!

rgds

ee