Analysis of the Stock Market: The fors and against arguments and discussion.

nesf

Right, I will start this and present my opinions on the pro side. Let's keep this clean and above the belt, and let's go into this acknowledging that most of the people who take sides in this argument will probably not change their minds about it, so lets keep it professional and objective ok

Hopefully this thread may help people who are undecided about analysis, or who do not know enough about it to make a decision, to make up their minds about this.


Ok, let’s start.

First off, lets look at the major accepted views that are held on the financial markets and the major models put forwards for understanding them or investing on them.

Now models are just that, models. They do not claim to be 100% accurate representations of the reality and they are at best accurate approximations. Now, bearing in mind this, I'm going to lay out some of my thinking on this.

The argument here is not, whether investing in the markets is a wise plan, but on whether analysis of the markets is a worthwhile endeavour. Now inside trading aside (i.e. the trading of shares in a company where you have insider knowledge, this is legal if done on a small scale, i.e. if you work in a company and decide to buy shares because the business is doing well, then this is legal).

I will keep technical jargon out of this, or more accurately I will seek to explain any jargon since this is an argument which may be of interest to people who don't have an academic or professional background in the fields.

I'll divide it up into sections so people can refute and respond easier.

What does technical and fundamental analysis actually involve (The simple version)

Fundamental analysis is analysis based on the fundamental data of a company (to keep it simple I'm not going to talk about the commodity markets here because that complicates matters a fair bit). Fundamental data is publicly available data such as annual revenue, declared profits, dividends per share, earnings per share etc. From this data one can see if a company is doing well financially. Based on these factors one makes out whether the stock in question is a buy or a sell (a very basic interpretation). A buy would be a stock with solid earnings, a good history of revenue and dividends, while a sell would be a company that is showing weakness or a downturn in their revenue figures etc.

Technical analysis is based purely on the charts. IE the stock/commodity graphs. It works based on the belief that certain chart patterns are repeated on charts, that looking at a chart can help one estimate where the stock is going, and through indicators do similar. This is a very simple definition, but I'm not writing a thesis here so I am not giving full definitions, google it if you want one.

What moves stock prices

There are many models for this and it is a very complicated topic (to 7, we can meet for a pint some night mate and have a long complex discussion about it ). Now stock prices move because of a few main reasons (I am not going to talk about commodities and the FX (currency), option, bond and forwards markets here because I want to keep it simple):

Company Data/Reports: Simply put, if a company returns good data then its stock will go up, bad data or data below estimates will do the converse. Also projected company data (as predicted by the top analysists in the industry) can affect a stock price a lot.

Sector Performance: Say it's a telecoms stock. The telecoms sector performance tends to influence each individual stock. If a market leader in telecoms posts a loss then the smaller stocks may drop in price even though they themselves have not posted a loss. Sector factors can cause a stock prices to be "soft", i.e. the prices are not reflecting the economic performance of the companies, but because the sector isn't doing very well people aren't willing to pay as much for a stock so the price can seem a bit low (It's actually a lot more complex than this but it's something of fractal style complexity).

Investor Sentiment: What the investors in the market (NYSE/whatever) are feeling like. Is it a bull(positive) or bear(negative) market? If it’s a bear market then people are loath to buy shares and well performing companies can actually have dropping share prices since there isn't anyone to buy them (stock prices fall on their own, but need buying for them to rise). Investor sentiment can be difficult to quantify, but looking at volume and open interest (not available in stocks), one can get an idea of it. IOW if a major market move (a big drop or rally) is accompanied by a spike in volume then one can take it as a solid indication of investor sentiment, one lacking said spike is artificial and the market will correct itself generally.

Political Factors: Policy decisions can have effects on stock prices, it's a complicated relationship and hard to quantify, but it can be clearly seen when one looks at market reactions to political news.

Economic Data: This is data like unemployment, monthly retail sales and the like. The release of major economic data has a profound effect. The markets react to these figures in varying ways. There generally is an expected value before each figure is released. The level of difference between the two will generally be related to the level of correction shown by the market. When an expected positive figure turns out to be majorily negative it can have very dramatic effects (look at the reaction of the US Dollar to figures in August last year for examples).

Psychology: This is a form of investor sentiment. People expect a stock to show resistance at a certain level for historical reasons, so when the stock approaches that level the number of bids (offers to buy) decrease as people wait to see what happens. Thus the charts show a levelling off near resistance levels.

Psychology has a very important role in analysis also; we'll come back to this later.

There are other factors, but I feel I've covered the main ones (I'm sick so I might have missed some).

Why fundamental analysis works

Fundamental analysis doesn't predict the future very well, what it does is allow you to see how a company is performing and compare this to its stock price. You can have a situation where the stock price is artificially high or low, due to whatever, while having fundamentals that would put the stock at a different price.

Fundamentals are all about solid economic data on the company. It allows you to evaluate the economic health of a company, and really is a solid tool to see if a stock price is all hype or based on real economic performance. It is a good tool for portfolio selection, but not good for short term trading. It is mostly used as a tool for the selection of long term share positions.

Fundamental analysis works, in that it does not predict anything, it just gives you a good picture of the economic health of a company. Now using predicted figures one can make predictions using fundamentals, but to be honest that is only as good as your predicted figures.

Why technical analysis works

Technical analysis or charting is a lot more grey than fundamental analysis. While some technical techniques are valid and proven to be effective (mechanical moving average trading systems out perform the markets, but not by much). Other techniques such as stochastic(a form of oscillator indicator) produce many whipsaws and false trading signals. But again, this is not a magic trick that tells you how to trade, indicators can be used to give an impression of where the market is likely to go, not where it will go. Big difference.

Charting also allows people to predict areas of support and resistance (psychologically important price areas which cause the market to react in a certain way). Also, sometimes market patterns emerge, such as double tops and such. Now personally I don’t try to pick tops and bottoms, I think that’s a loser’s game. Trading against the trend is never a good idea, m’kay.

Charting’s merits as an analysis technique really depends on one’s perspective. Personally my opinion of its validity is this. It is a self fulfilling prophesy, in that if enough people use it, and trade on its predictions then its predictions will be true. As a tool then it does have some merits. Its practices are more than a little obscure and can at times seem more like astrology applied to the markets rather than a scientific technique.

However, and I can’t stress this enough, some of its more basic tenets run true. A trend is more likely to continue than it is to change, there is a lot of statistical evidence to back this, last time I say figures it was around 70% for the chance of a monthly trend to follow into the next month. Now this is all well and good, but it does not mean that you can just magically create profits out of nowhere. Trend following works very well, when the market is trending. But the market spends only about half the time in an actual trend, so one needs to be able to tell if the market is trending or not.

This is where the quality of the Analysist and not the technique comes into play. A good Analysist who is familiar with charting knows when to apply each tool and when to trade and when not to trade. This is the art of it. There is a random element to market movements, I don’t think anyone who has worked with the markets can honestly say there isn’t. Thus the element of chance does pervade all attempts at analysis.

But the psychological groundings of charting techniques can not be under-estimated. Technical analysists are used and employed by all the major players in the markets, so dismissing the techniques out of hand really does not make sense. Whether or not you believe that the gains made through charting (minimal to be honest, but charting does increase returns) are worth it for the amount of effort put it, is a matter of opinion.

nesf

Market Analysis is it worth it?

Right this is a tough question. I think personally that it is. But that is only my opinion. When I decide to do a little analysis on here I do it in good faith. I present what my analysis is with only a view to providing information for people who want it. I back up my research by saying exactly why I think the way I do. If I merely went buy this stock, it looks good, then I should be shot down by a barrage of comments. But when I do offer analysis I will give specific references to why I believe certain things. I’ll say where I expect resistance and say what the MACD or whatever indicator I want to use is saying.

Analysis at its worst can be just ignored. But to dismiss it out of hand because of a belief that it is impossible to analyse a stock or commodity is not constructive. Yes, I agree that analysis at best is an educated guess. I will never come on here and say that my predictions are 100% accurate or that my analysis is fool proof. It isn’t.

For example, in futures trading, your average trader is correct on his analysis around 55% of the time (this is for short term trades btw). The important part of analysis is, that it can aid in the timing of entry and exit of trades. If you are trading for the short term, then you are in for a rough ride my friend, short term trading is nasty at best.

Now long term investments are a completely different story. Analysis is key in the selection of shares. Now the important point here is that one needs to ask the right questions. Will a stock go up or down? That’s not the right question. Does the stock in question have the potential for strong gains? Now that is far more answerable. A stock's volatility, past performance, sector performance and prevailing economic conditions are all things we can factor into our analysis.

When I say that a share is a nice one for a punt, what I am saying is that this stock has in my opinion a good chance of gaining on its present price. That is the best I can do. I cannot say how it will behave over the next 12 months or what value it will be. I can say what I expect it to do, and why I expect that. That is the best I can offer. People can take it or leave it, I don’t mind.


Please support the Investments/Financial Markets/Money Board proposal if you would like to read more like this.

Link: http://www.boards.ie/vbulletin/showthread.php?t=246825

Thanks for reading, I hope the above was informative. I am stuck in tonight with a chest infection, otherwise I'd be in the pub right now. I am sick, and my thinking is more than a little fuzzy at the moment, so please excuse me for any inconsistencies above. If there any mistakes please outline them in a post and i'll edit the above.

Edit, damn I've over stepped the mark, so I'll split this into two posts.

7mountpleasant

Well put argument Nesf now I will give my tuppence. My main reason for disagreeing with the validity of technical and fundamental analysis is simply the fact that nonody has ever been proven to profit from it. Despite claims to the contrary , your average active fund manager does not make a risk adjusted return which outperforms throwing all your money into an Index tracker. The simple reasoning behind it is this , each transaction involves two parties. To consistenlty generate abnormal profit you must long or short an asset depending on your view of over/under valuation. However the key point of note here is that in every transaction there is a counterparty who has the complete opposite viewpoint of you. To suggeest that on a consistent basis you can find a counterparty who will sell you an asset which is undervalued or buy an asset that is overvalued does not hold in a market of many many participants each trading upon the same information. This is not to say that it is not possible to make a profit from the purchase and sale of derivitaves/shares etc but simpy that this profit is the return for taking on board the risk of holding that security. If we return once again to my friends the fund management community the funds which consistently return the best average risk adjusted returns are not the Active/Momentum or any other type of active manager but simply your Normal Index tracker such as the Vanguard S&P 500.

nesf

7mountpleasant wrote:

Well put argument Nesf now I will give my tuppence. My main reason for disagreeing with the validity of technical and fundamental analysis is simply the fact that nonody has ever been proven to profit from it. Despite claims to the contrary , your average active fund manager does not make a risk adjusted return which outperforms throwing all your money into an Index tracker. The simple reasoning behind it is this , each transaction involves two parties. To consistenlty generate abnormal profit you must long or short an asset depending on your view of over/under valuation. However the key point of note here is that in every transaction there is a counterparty who has the complete opposite viewpoint of you. To suggeest that on a consistent basis you can find a counterparty who will sell you an asset which is undervalued or buy an asset that is overvalued does not hold in a market of many many participants each trading upon the same information. This is not to say that it is not possible to make a profit from the purchase and sale of derivitaves/shares etc but simpy that this profit is the return for taking on board the risk of holding that security. If we return once again to my friends the fund management community the funds which consistently return the best average risk adjusted returns are not the Active/Momentum or any other type of active manager but simply your Normal Index tracker such as the Vanguard S&P 500.

I disagree, well managed funds do well above market returns, look at Merryl Lynch (sp?) for example.

The concept of the profit being a "reward" for the risk taken is a very important factor. I'm sure you are familiar with Portfolio Management (although it's more of a financial mathematics topic rather than a finance one strictly), where the return is balanced against the volatility of the stocks chosen. But I digress.

The managing of funds is big big business. I don't think anyone who's reading this thread would disagree. Risk adjusted returns are a very complex issue though. Does risk adjustment really reflect the validity of a trading strategy?

The question is quite complex. Investing money in a fund/stocks/whatever does not work like a bank. You do not get a guaranteed return on your money. You can't annualise a return and average it across an industry and expect to have a figure that actually means anything. Statistics are great, but as i'm sure you know, they can be manipulated in alot of suprising ways. For instance, 10% annual returns from a managed future fund is not unheard of. Hedge funds for instance produce similar returns in a trending market. The key word there is trending.

An index following fund is a low risk option and low risk options always return better results when we adjust for risk (or at least nearly always). So quoting risk adjusted figures means little since you are not quoting actual returns. They are very valid figures but I feel that you are misusing them. Risk adjusted figures are a great way to represent specific investment options when the money needs to be protected, ie in a pension fund for instance, or a situation where someone wants to "make their money work for them" but where they can't afford to lose it.

If however one looks at the returns without adjusting them for risk, then it is a very different picture. Depending on the market traded (futures for instance are notorious for high returns combined with very high risk to match), we get different projected returns.

The key thing to remember here is, if one is willing to take on a high risk, then yes one can reap a big profit. You could also lose your investment. As you said profit is a reward for risk. There is no way to trade on the financial markets and make a profit without risk, unless you are lucky and skilled enough to discover an arbitrage situation. These situations do occur, but modern banks have alot of people trained in spotting them whose only job is to spot these arbitrage situations and take them (ie this means that there is a window realistically of minutes (a half hour tops) for arbitrage on the financial markets). So for all intents and purposes we can assume that arbitrage on the modern markets is non-existent (not actually true, but it's easier to work if we assume it is).

Arbitrage is riskless profit which financially is a contradiction in terms. The best example I can think of is from the bookmaking industry a few years back. At one point in the US elections it was possible to back Bush to win the election at odds of 3 to 1 in the US, and back Gore to win at odds of 3 to 1 in the UK. Now if I (and I did) put bets for both Bush and Gore, I am recieving a profit regardless of who wins. Therefore I have no risk, and am guaranteed the profits. This is what is called an arbitrage situation. As an aside arbitrage betting is an interesting way of making money these days With the distribution of prices between bookmakers and sites like Betfair, one can with a fair bit of work, find these arbitrage situations day in, day out and actually make some nice money doing it. But thats a topic for another thread.

Now, when we look at the managed funds versus index following funds we can see a clear relationship between risk and profit. Now I don't need to be stating this for you, I'm positive you are totally aware of all this, but I want to keep this as well explained as I can so that people not trained in the area can follow it. If the two of us just fell into using jargon then while yes the posts would be alot shorter, we'd end up with something unreadable to anyone who just has a general or non-technical interest in the area. If I'm coming across as being condescending please excuse me, it's not meant in that way.

Eh got to go see the hitchhikers guide to the galaxy now, um I'll finish what I was saying when I get back. Damn stupid real life getting in the way!

7mountpleasant

Maybe I digressed myself in my reply. But when I spoke of Risk Adjusted return and the lack of abnormal risk adjusted returns for fund managers (essentially a very interesting concept known as Jenson's alpha) I spoke of a return over and above the return associated with the level of risk. Now Here I will get a little bit technical when the returns of Managed Funds are regressed back over a time period of usually five years and controlling for things such as survivorship bias (i.e. looking at only the winners) only an extremely small percentage have ever produced abnormal returns persistently. While I do agree that the performance of hedge funds has led many to question the so called efficient markets question this can be explained by two things first the level of risk taken on board by Hedge funds is extremely large (for evidence look at the famous long term capital management fund which nearly caused a major banking crisis in the U.S) and secondly that they were able to take advantage of cracks in the system (mainly bond rated below investment grade). But look through the financial pages any day of the week and you will hear of both wild gains and losses in hedge fund which confirms in my view the status of a lot of hedge funds as financial poker players.

Now I know I digressed an awful lot there so I will return. The reason why I concentrate so much on Managed funds to explain my opinion is that the buy side analysts who are the main practitioners of such analysis, are the people who run/advise these funds therefore is such analysis were able to return abnormal returns is here we should see it. The fact that we don't in my opinion is an indictment of such analysis.

nesf

Good discussion, sorry for the late reply, spent most of today recovering from my hangover.

Ok lets, outline a few key facts we can agree on.

1. Profit is proportional to risk undertaken. This must always be understood and appreciated. Riskless profit doesn't exist in order to make the equations simplier, but we know that it does exist, but only in very small quantities for short periods of time, so we can assume that it is negligible when measured against risk driven profits.

2. The benchmark return rate (an index rate or similar), can and will quite regularily be exceeded by a managed fund. I don't think an averaged figure is representitive of these funds though, simply because it isn't a regular distribution and a non weighted mean means little to such data.

3. We agree that there are people out there who are far too sucessful as traders/investors. Warren Buffet is a good example. These people's existence is a clear indication of the problems of the efficient market hypothesis. (I'll do a refutation of the EMH in a while).

4. There is no simple answers for investing. Simply analysing a market is not a fool proof method to sucess. At the best of times you might have better than 50% odds on being right about a short term prediction, statistically 60% is about the best. The reason why you make money is because you learn very quickly to cut your losing positions quickly and leave your winning trades run. This is clear cut for short term investments, but not so much for long term investing. A temporary drop in a share price due to bad publicity might not be a clear sell signal. Shares have an intrinsic value (related to earnings per share etc) and the market will generally correct itself to this value. A share can be overvalued or undervalued, but they rarely stay in such a position for long. Also, shares pay dividends, dividends are nice, they are essentially free money to an investor.

5. Analysis is built into the markets. Since the creation of formal exchanges, people have attempted to predict market movements. This has resulted in psychological factors being a huge factor in market movement. The markets move for a variety of reasons, one of these reasons is that people believe in their analysis.

A simple example. Stock A is trading at $56.65 and is falling. It has fallen over $10 in the past month. The market consists of three types of traders, bidders, sellers and neutrals. At any one time the majority of market participants are generally neutral unless the market is trending very strongly. Now the key thing to remember is that stocks can fall on their own with little or no trading being undertaken, but for a stock to rise the bidders need to outnumber the sellers (at least in a figurative sense). However there is a well known resistance area at $52.00. Now if this resistance area is very well known, well tested and most importantly believed in the following can happen.

As the price of Stock A approaches this resistance level, more and more people turn from being neutral to being bidders. They want to buy this stock because they believe from their analysis that it will bounce up from this resistance level. Now as the Stock approaches this resistance level, the number of bidders increase. Now what can happen, if there is enough belief in this resistance level is that the stock can start to rise again. As it rises more of the neutrals become bidders since they think the stock price has hit a turning point and is going to at least retrace part of it's fall. Time to buy cheap as they say.

Now what happens next is quite important. If there is enough belief in the stock, it may continue to rise and rise, and this bounce from resistance could start a new trend.

Whats more likely is that after the stock has gained maybe a dollar or two of ground, people will start profit taking. Alot of people who bought at the resistance level might not have faith in this stock to keep on going so they are taking their profits and going to applaud themselves for a job well done and a resistance level well spotted.

Now say if the market falls again, and retests the resistance again. What will happen? It's a clear rule in charting that a resistance level that has been tested many times is much stronger than one that has only been tested once. If the market continually tests a resistance level, people might start to believe that the stock really should fall and is not worth it's present price. If that happens we can have a breakthrough, now when a resistance level is broken, you will generally see a volume spike. People are loath to trade a stock 10 ticks off a resistance level so there will be little action as people wait for the reaction when it hits the resistance (ie mostly neutral market participants). Now if they decide that the resistance isn't going to hold you will see a huge spike of volume at the reisistance level as people follow the herd and go bearish on the stock and suddenly want to sell it. People with long term positions will liquidate them if the reisistance isn't holding, and neutrals will want to short the share if they can. You will see a very sharp drop in price.

Now the above example was just me trying to get across one of the simpliest psychological effects that affect the market. These effects are predictable and do happen. The argument that results from this is that technical analysis is a self-fulfilling prophesy. If enough people believe in it, it will happen. Although to be fair resistance levels can be found in even the earliest charts, so it can be argued that psychologically important price levels are going to exist in any market that involves people. People like certain numbers, if a stock price reaches the level of it's last major low of the past 5 years, then people start thinking, "well it recovered from this point 5 years ago, maybe now would be a good time to buy."

That there is a random unpredictable element to the markets is without a doubt. TA can explain some movements, but most of the time at best it can only guesstimate whats actually happening. TA is basically an exercise in mass psychology. Mass Psychology at the best of times is very hard to predict. It is possible to predict it in a general sense, but there will be times when your predictions and what's happening will be totally distinct and bear no resembelence to each other.

This I suppose is where my opinion enters the fray. Technical Analysis just like Fundamental Analysis works very well at certain times in certain types of markets. If a market is trendless and is trading in a channel, then TA is of limited use to a cautious trader. An aggressive trader may try to "trade the channel" but I personally wouldn't be putting much money behind it. Now if a market is trending (which it is around 66% of the time) then TA is a god send, and the simple techniques such as moving averages, MACD's etc can be used to great effect. The problem arrises in that you need to know when to be looking at the TA or not. This is a matter of opinion, and really is the sign of a better analysist. Use the right tools for the right jobs, and you'll do well. Try and screw in a bolt with a hammer and you are heading for trouble.

The argument over the validity of analysis techniques is a long and complex one (as can be seen thusfar). I'll continue it in my next post

nesf

Now to continue

My personal views on analysis are this. Analysis is a valuable tool when used correctly. It does not predict the future accurately, but it is alot more efficient than trading blind.

The reasons why huge returns are not posted by the big funds are as follows.

Risk = Profit. This is the immutable law of the world. To make a large profit you need to take large risks. Large risks are not the way most large funds are managed. Large risks are often taken by people invests 2 or 3 grand in the hope it'll turn into 20 or 30 grand. They don't care about losing the original capital because there is a half descent chance that they might make a killing. Now the thing is, fund managers don't think this way. In general. Most fund managers take calculated risks, carefully balancing risk versus profit but AT ALL TIMES preserving their funds. Fund management is crucial in trading. Generally if you are trading from an account, a third party carefully monitors your loss rates and if you lose too much too quickly you will be ability to trade will be rescinded and you will be put on a months duty elsewhere to allow yourself to get things together. Fund Managers can't take on the risks that turn large profits (80%+ returns) simply because if they don't win they will not be able to trade afterwards. Fund Managers can only take on risks that their fund can absorb. For instance a general rule of fund management is that at no time does a single trade risk more than 3% of the fund. This is rigidly enforced. Trading at 3% of the fund makes it very difficult to achieve high return rates over a sustained amount of time, but it also will prevent you crashing and burning very quickly. Yes the occasional big win will happen, but fund management is all about sustained return not big one offs.

Now these limits are there for very good reasons. Nobody wants to see the fund go bust, and this could happen if a fund manager was left to take on any risks he liked.

The above is obviously going to limit the return rate from a managed fund but it also decreases the risk of a fund going bust dramatically. Remember it's 3% of the fund, not the original fund, so as the fund gets smaller, the fiscal acceptable risk gets smaller also.

nesf

Now a caveat to the above risk = profit equation is this. It is quite simple to make a large financial profit when you have a very large fund to work from. But the rate of return (ie the profit expressed as a percentage increase of the principle sum invested), is not large.

The problem with alot of the more efficient techniques of analysis is as follows:

The most effective mechanical trading system is a very simple one consisting of 2 or more moving averages, and trading signals are generated when these two averages cross each other. Whipsaws can be reduced by adding a third average. Now the time period chosen for these averages allows us to choose the type of movements we want to catch. For instance, a 200 day and 400 day moving average combination will only be effected by the biggest trends. A trend would have to be going on for at least 6 months before the two averages generated a sure signal in it's direction, but you could be sure that you'd only get signals from big trends and movements. However you'd be trading once, maybe twice a year. Not a way to make a living.

If however we chose the industry standard 7, 14, 21 day distribution we'd pick up any weekly or monthly trends. Now if we set up a mechanical trading system that just followed these trading signals generated by the above 3 we would get an average rate of return of significantly more than the benchmark.

The problem is, that during the time this mechanical system traded, you would need very deep pockets just to stay in the game because by blindly following the trading signals areas of huge loss do occur. The average return is brilliant, but this system has a very high probability of burning out the fund it's trading from at some point, because there are times (like in a trendless market) where moving averages should be ignored! But the system can't know that. There is no easy way to tell if a market is trending or not until about a month or two into a new trend. But by that point you've lost quite a few profits because you weren't sure if the market was trending or if it was still trendless and this was just a temporary movement.

Mechanical systems which are more complex than this actually produce a lower rate of return!! The problem is that the more complex a system becomes the more likely the system will be wrong. The markets are ruthless, no amount of complex trading rules will benifit you. Simple approaches are the best when mechanical systems are concerned.

Now the problem is, when the human element is added in. People make mistakes and can convince themselves of things that aren't true. We've all seen that. Now in the financial markets emotional attachment or ego tied to a position can be crippling. If you cannot admit you are wrong and cut a position without a second thought, then you really have no place trading. Trading is ruthless, it doesn't matter how good you are, you will have losing streaks, and if you can't get over them or weather the storm, then you will fail.

Now the above example highlights a very important fact. The Efficient Market hypothesis is wrong. It's not a bad theory, but it simply isn't true for the financial markets. I do not believe that there is no random element to the markets and that they are predictable all of the time. However, from experience and factual evidence, it is clear that the markets are not 100% random. That will be the subject of my next post

nesf

Right the Efficient Market Hypothesis then, the major enemy to analysis

Excellent info available here: http://en.wikipedia.org/wiki/Efficient_market_hypothesis

The definition that we will work off of is the following:

The hypothesis that the stock market is a highly efficient pricing mechanism. Efficiency in this context does not refer to the organizational and operational aspects of the market or to the efficient allocation or resources within the economy but to the capacity of the market to convert information into shares prices. There are three formus of the hypothesis: (a) weak efficiency: share prices move independently of previous movements. (b) Semi-strong efficiency: shares prices respond instantaneously and in an unbiased manner to all information publicly available. (c) Strong efficiency: share prices fully reflect all relevant information including that not publicly

Source: Google Definition

Now for any person working in the industry, it is accepted generally that the efficient market hypothesis is just wrong. It is a reality that the markets do react to historical data, my reasoning behind this is because of mass psychology. Believe when I say that anyone who has worked with the markets will tell you that they are not totally random. The stock market crashes of 1979 are the clearest example of this, but you can find examples of it in almost any chart you care to look at.

However.

The efficient market hypothesis is not a bad theory. My academic background is quite theory heavy so I've alot of respect for well worked tidy theories that approximate the situation it's trying to model well. The problem with the efficient market hypothesis is that it isn't a very good approximation when you're looking at the markets over the long term. Overe the course of 5 minutes the markets might exhibit totally random behaviour, but over the course of 5 years they show anything but a totally random nature. So what are we to do? It's a nice theory, and some of it's predictions are true, if not all of them. The distribution of fund performance is not regular or fitting with the predictions of this model.

But a problem with the model (like many theories) is that while it works excellently in an academic sense, it doesn't hold up well when one goes into the real world. It is taught by many colleges as a key part of understanding the financial markets, but that is generally because it is a very easy model to teach when compared to the likes of behavioural finance.

The efficient market hypothesis would mean that analysis could never work, and it would be sheer luck if any of the predictions were true. However the reality is quite different.

I'm going to propose something a little different, and I'll throw it open to interpretation at that point.

What I'm going to do is formulate a simple model which is representitive of the markets we know and love, and which can explain quite alot of what goes on.

First, we will assume that at least some of the time, the price movements of a financial instrument will be influenced by a selection of the following: historic price data, psychological factors, economic data, news and finally totally random movement.

Now how much of the market is random fluctuations and how much of it is cause driven? I'm not going to put figures on this, I don't think my guesses would be accurate. However, what I do want to reinforce is:

The markets move due to a variety of factors, but not all of them can be accounted for by analysis. There are major factors in the movement of the markets which are for all intents and purposes random and without predicatable cause.

But as a counter-point, we will also assume that there are other major factors in market movement which are not random. These factors are impossible to know in their entirety, but we can to some extent predict them.

Now I see analysis as being a valuable tool in assertaining at least some of the forces behind market movements. But the important thing to realise is, that there are random forces effecting the markets which are not predictable.

What I propose is, that the markets are quasi-efficient. That is there is a large percentage of the market movements which are random (read: efficient), but there are equally large factors which are inefficient (ie predictable to some degree).



I'll take this concept forwards to my next post then I'll be finished for tonight

Hope this is of interest to some of yee

nesf

After reading the above post something occured to me, I did not explain what I meant by random movement.

Efficient movement basically for all intents and purposes means random movement. This is because an assumpton of the theory is that news that will effect the share is essentially random in distribution and not predictable. Thusly the movements are random.

This is not the same as the random walk hypothesis. The randomness of the EMH comes from the random nature of factors that influence the stock rather than the random movement of the price itself. We can consider the movement to be random since the basis for the movement is random, but it is not the same as a truly randomised movement. Things like inflation will make a stock rise in price over time without any actual change in the "actual value" of the stock.

But a true discussion on the various types of randomness and chaos is far beyond the scope of this argument.


Right, ok. I think I've made a fair argument to back up analysis. I just want to touch upon a few key topics before I go to bed.

First off. While analysis does allow us a small degree of foresight with regard to stocks, this is tempered by a few factors. Firstly the random element of the market will tend to cause enough fluctuations to nullify quite alot of analysis in the extreme short term. As we go over larger and larger time scales this random nature of the market is smoothed out and becomes less and less of a factor, but as we stretch out the timescale our analysis becomes less and less accurate. This is the balancing factor in my opinion and is what prevents extreme returns from the financial markets.

Analysis itself isn't flawed, it is quite good at predicting what is predictable about the markets. For example through fundamental analysis we could discover a stock whose share price is much lower than it's asset per share ratio (for instance, there are other factors, but lets keep things simple) (asset per share ratio is the total physical assets of a company divided by the number of shares, ie the actual real value of a share if the company was liquidated right now without having creditors). Now the share price usually is more than this asset price, since this asset price per share is a kind of intrinsic minimum value of the share. Now, this stock will most likely rise, as the market balances out to a price closer to this asset/share ratio. (If however the company is showing heavy losses in the past few years then it might be heavily and debt, and the low price might be justified. This is why you research a company very closely before committing a lot of money in it). The market tends to balance in this way, in both directions. This is predicatable movement. However fundamental analysis cannot tell us when this price rise will happen. The market could balance out yes, but it might be in 5 years or 5 months. We have no way of knowing when this balancing will occur. Market timing is an art, knowing when a share will move is such a complex concept that few traders ever really get it (myself included).

These are just my thoughts and yes I've thought about this quite alot over the past few years that I've been involved with the markets. When I started I was a true believer in analysis. That is most definitely not the case. I've seen the random unpredictable side to the markets. I've also seen them act exactly like I predicted, and this happens a lot more often than pure chance would have us believe is possible. I like to analyse things and model things. I've given alot of thought and researched this topic extensively. I've seen the evidence for both sides, and drawn my own conclusions.

Analysis is without doubt in my mind, a useful tool in the arsenal of the investor, but it is not a simple path to wealth. The more volatile a stock, the more potential profit is achievable, however a highly volatile stock will tend to be one that seems to display manic depression. It will reach dizzying highs on the slightest bit of good news, or come crashing down with the slightest hint of negativity. Also the random elements will be stronger in such a stock generally. A slow stable stock doesn't show major random movement, however it will also not yield large profits since the share value will be very efficient and not need to vary much.

My thoughts run along the lines of alot of thinking in behavioural finance (wikipedia article) but are mostly from my practical experiences as a trader and analysist combined with theories from finance, financial mathematics and economics. The above is pretty much my thoughts on the matter from both a theoretical and practical orientated viewpoint.

I didn't touch a lot of topics in the above, such as risk avoidance, because I am not writing a thesis here. This is just my reasoning behind why analysis is of some value with regard to the markets. The fact that analysis is limited by certain factors is indisputable. But if the human element was removed from the stock market, I believe that alot of why analysis works would leave, and only the simpliest techniques could survive (like the moving average). Occam's Razor is of particular importance in the markets, the more complicated your approach the more likely it is that you will fail.

Now the above does not mean that complex indicators do not have their place. Complex indicators are generally measuring quite simple forces, and are only complex because of refinement rather than complexity in concept. One of the best examples of this is cycle/wave theory (I'm not going to get into it here, it's a damn complex topic). Basically it states that the markets move in waves, like a bull cycle followed by a bear cycle. Now the theory goes, that there is many many cycles at work in the markets and that they are superimposed upon each other. Superimposing cycles is a very good way at making something look random btw.

Now the key concept behind wave theory is very simple. The implementation though is extrememly complex. Picking out an individual wave from the noise of the market is no mean feat. There are hundreds of examples of cycles and waves in the markets (and in life), and yes it's possible to spot these cycles after they have happened. But to pick them out while they going on is very difficult and complex. There are a few very talented people who achieve very impressive and consistent returns from wave analysis, but they are the few. For most analysists wave theory is just too damn complex to be of any use.

In conclusion, I suppose what I'm saying is that analysis is good at what it does, but alot of what happens in the market is not predictable, and thusly why analysis can't predict the future exactly. But conversly it does explain why analysis can be so right. It's right far more often than the efficient market hypothesis would allow. If one looks at the markets as being only partially predictable and approach investing with this in mind then one can achieve good returns, nothing stellar, but better than would be got through the banks.

And hell, it's damn interesting anyways!

I hope my musings on this were of interest.

7mountpleasant

Will write up a full reply later in the week. However one point which I will talk about is the ascertain that EMH is wrong. I have heard the what about Warren Buffet , Peter Lynch argument many many times. At first glance these two cases seem to point to the failure of such efficient markets, however deeper analysis of both reveal some interesting facts. Firstly on the Peter Lynch side, while the performance of Megellan was generally accepted to have beaten the market when Lynch relinquished his stewardship of the fund the composition of it was almost identical to the market. As lynch himself alluded to the reason for his outstanding success was the surprising and abnormal growth in a number of Franchise Business's (most notably home Depot) when analysing Buffet one notices similar trends (also a major holder of Home Depot). In essence Wiltshire is essentially a Hedge Fund Manager and the wild swings of his portfolio will point to the fact. Also as well as this Buffet has made large profits from certain markets which in my opinion have been at times inefficient (most notably Debt rated below Investment Grade thereby making it impossible for American Financial Institutions to purchase it due to Capital Requirements). The existence of Buffet and others like Carlos Slim does not in my opinion point to inefficient Equity Markets. In fact their success has been counterproductive with the value of funds under management in hedge funds last year rising to $1000bn suitable opportunities will be impossible to come by for these investors in the future.

The reason market participants state that they believe that believe that the markets are inefficient is because if the markets were efficient then the noise trading which they participate in and the recommendations of sell analysts whose remuneration is linked to trading volume and not economic sense would be futile. It is the increasing frustration of people outside the Fund management community with the poor returns and large 'management fees' that has led to the spectacular success of such funds as the Vanguard 500. In fact even last week a new fund was launched by the ISE known as the ETF in response to the growing demand for index tracking funds. While I admit that the large volatility of certain securities prices could leave people to believe that technical and fundamental analysis could be a valid investment strategy ,the overriding evidence points to the fact that share prices follow a random walk and technical analysis attempts to reduce to mathematical equations what is essentially a human transaction. If you were able to identify a recurring trend in a stock price the profit which could be made from peaking and troughing would be enormous. However the fundamental point still remains the price which appears on your screen is not the result of some magical algorithm but simply the price someone is willing to buy or sell that asset for. While the mechanics of settlement and information may have changed that fundamental fact has not.

nesf

I'll hold off on reply to give you a chance to formulate a full rebuttal.

Going to the cork boards meet this friday?