CAPM (Capital Asset Pricing Model)

Pepper

Junior member
Messages
10
Likes
0
Capm

CAPM. Does it help to forecast stock market?

There are so many discussions about this model. So many people argue how effective this model could be. The discussions about it have been going on over the entire lifetime of the model - from the very moment of its development. So many people predict the end of the model, but in spite of this fact the model is still alive.
In this article we’ll point out some CAPM features and try to find answer to this question.
First of all, let stop on definition. CAPM abbreviation means Capital Asset Pricing Model. Capital Asset Pricing Model (CAPM) is an equilibrium model that estimates the return of financial assets. This regression model compares the risk of exact asset and stock market. CAPM states on assumption that investors make their investment decisions base on risk and return. CAPM is a well-known model; companies such as Merrill Lynch use it.
This model was developed by Nobel Prize winner William Sharpe, John Lintner, Jack Trainor, and Jan Mossin.
This model they divided risk on systematic and non-systematic. Systematic risk stipulated by economic and stock market conditions that affect all market securities. Non-systematic risk depends on the concrete issuer.
It’s important to mention that it’s impossible to decrease the systematic risk. We can estimate the influence of stock market to exact security. As a measure of systematic risk CAPM use beta coefficient.
It’s possible to influence on non-systematic risk. Non-systematic risk can be decreased by diversification on stock portfolio. Portfolio should consist of different assets. Stocks should be originated from different sectors of economy and have different prices.
There are two important CAPM parameters describing a concrete asset, are:
· "Beta" parameter. Beta is an asset market risk parameter. Beta is the ratio of the average stock yield relative to the overall market portfolio or a stock market index such as the S&P500. It is calculated using historical data taken over a year or more. Once beta is calculated, it is considered to be a predictor of future market behavior. If the stock market goes up (or down) by a particular percentage, the theory is that there is a tendency for the stock itself to go up (or down) by the same percentage multiplied by beta. Stocks with a beta greater than 1 are considered riskier; when the stock market fluctuates, the high-beta stocks fluctuate even more. The more is beta the riskier the stock. The exact calculations are necessary for managers to select the assets that best fit their investment strategy. Managers can form different types of portfolios, conservative, aggressive, balanced etc. according to beta coefficient.
· E parameter corresponds to "residual" yield dependent on specificity of a concrete asset. It is matched with non-systematic risk, which could be reduced by creation of asset portfolio.

CAPM says that if you know a security's beta then you know the value of return that investors expect it to have.





Ep portfolio

r
0 beta 1
r is the rate of a "risk-free" investment, i.e. cash;
Ep is the return rate of the appropriate portfolio.

CAPM, as any other model, has its area of applicability. This area is defined by simplifications, accepted during model creation. Such assumptions filter out unnecessary complexities and allow effective application of mathematical core, which translates empirical observations from area of sensation to the area of knowledge. Therefore, when creating CAPM, the following assumptions have been made:
· Investor is guided by only two factors - yield and risk;
· Investors operate rationally - with the same expected yield they prefer an asset with the minimal risk
· All investors have the same investment timeframe;
· Investors evaluate asset key parameters in identical manner;
· Individual investor behavior does not influence equal prices of an asset;
· There are no operational costs or hindrances, preventing free supply and demand of assets.
Let’s take a look on CAPM figure

CAPM model.
Beta an asset market risk parameter represents straight-line inclination degree.
E is average "residual" yield, describing an average asset yield deviation from "fair" yield as shown by the central line.

To create a graphic interpretation of the model, plot on a plane points, whose horizontal coordinates represent a market portfolio yield, while the vertical ones are the appropriate asset yield. One of the major stock indices, for example S&P500, is taken as market portfolio. Viewing the formed cloud of points closely, one can note that it is extended along some straight line - that of characteristic line of the given security.
The main assertion of the model is that price yield of a selected asset (or portfolio) is directly proportional to price yield of the market portfolio.
As were mentioned before, the 2 main parameters are Beta coefficient and, describing angle of straight line inclination and E parameter, describing degree of cloud concentration along the straight line.
"Beta" is a parameter of asset sensitivity to changes in market portfolio price. If, for example, "beta" is equal to 1.5, it means that when the market portfolio changes by +1%, the asset price will change by +1.5%. Assets that are more sensitive to the market are matched with greater "beta" values. This parameter is responsible for systematic (or market) asset risk, which is impossible to diversify.
To analyze stocks traded at NYSE, NASDAQ, and AMEX you can use products based on CAPM model. Such products are executed at top mathematical level. They can help you to calculated values of "beta" parameter, estimate when asset overvaluated or undervaluated and carry out asset yield forecast based on market portfolio yield.
Maybe CAPM model is not the best one but it can help traders and investors a lot!

For full text version click on http://www.e-mastertrade.com/en/main/reviews/default.asp
 
CAPM is worthless for trading as it is longer-term in nature.

Even in the long-term it's value is dubious for financial markets as it is based on a linear model when it is widely understood that market returns are non-linear in nature.
 
Rhody Trader said:
CAPM is worthless for trading as it is longer-term in nature.

Even in the long-term it's value is dubious for financial markets as it is based on a linear model when it is widely understood that market returns are non-linear in nature.

This really depends on your trading time horizon, one could easily argue that "trading" could involve holding longer term positions. It's also worth noting that CAPM and variants (such as Fama-French three factor model) are currently part of a University finance education if you take any classes that deal with portfolio management. Chances are also pretty good that many professional fund managers are utilizing CAPM concepts as an aspect of their decision making.

Can you suggest a better model?
 
I learned CAPM at university myself, but even the Finance department chairman of the time readily admits today that such models are outdated and the modern financial theory has moved on. I'm not sure exactly to what as I've been out of academia for quite some time (though I understand the analysis of high frequency data is hot), but I have not often been impressed with the practical nature of much research that is produced, at least from the perspective of an individual trader. More so, perhaps, for institutions. The assumptions that go in to these models are not realistics in so many cases ("Assuming no transaction costs ....", "Assuming constant volatility....", etc.)

As for CAPM specifically, there is value in it conceptually. Understanding that some securities will be more or less volatile than the general market is handy. The problem comes in application. Betas are of dubious real worth since they are in flux. Also, how do you measure the market, both for determining beta and for determining the market rate of return? When any real measure of precision is desired, the model falls apart. It's a framework for making a guess.
 
Rhody Trader said:
Also, how do you measure the market, both for determining beta and for determining the market rate of return? When any real measure of precision is desired, the model falls apart. It's a framework for making a guess.

I believe that fund managers adapt the CAPM concept to be index based, i.e. looking at the beta of a given stock in relation to, say, the S&P 500. By using a multi-factor variant of the CAPM, one could compare the historical performance of a stock to inflation, oil prices, etc., which could help with decision making - i.e. "What might this stock do assuming high inflation in the coming year". Because you can calculate/estimate a beta using regression, I think that CAPM is in fact quite valid.

All of the above said, CAPM is likely overkill and/or not particularly usefull for a day trader.
 
CAPM is a very good theoretical model, I believe the Nobel prize goes to show this. However, it misses one crucial point. It's validity rests on the assumption that the markets are rational. It uses this assumption because without it, the model simply could not exist. The markets are definitely not rational though. In the long term rationality does seem to improve, however, which is why people often claim that in the long term CAPM retains some practical validity, however random walk and efficient market theory tell us that it is impossible to predict the future movement of a stock or index as it is sensitive to news which is as yet unknown, therefore, while CAPM would retain validity in the long term as irrationality is averaged out, random walk tells us that it is impossible to determine the state of the markets at any point in the future, from a fraction of a second to a decade, so the conditions we would use in our calculation of beta etc in using CAPM are only valid at the time at which we make them and become invalid immediately after. So while CAPM is a very clever model, and I could only hope to come up with something to rival it (clearly I can not). It is not valid as a practical tool any more than any other type of technical or fundamental measure, and due to its inflexibility, sometimes less so.
 
danfreek said:
... the model simply could not exist. The markets are definitely not rational though...., however random walk and efficient market theory tell us that it is impossible to predict the future movement ....

Everything you were saying was great right up to the point where you mentioned the Efficient Market Theory. Recall that the EMT also depends on rationality, so when you use that to attack CAPM, your argument is contradictory.

Don't get me wrong! I'm agreeing with your main point about CAPM.

I'm not a big fan of Random Walk, though. Most of the reasoning I hear defending it involves something having to do with some kind of random series, when plotted, looks just like the chart of a traded instrument. That essentially says that if something looks like something else, it must be that other thing. This is not a good defense. It's nearly impossible to prove randomness, of course, but let's not say that just because the markets look random that they must be.
 
that's true the efficient market hypothesis does rely on rationality (which is why I don't pay attention to it) but it is in direct contradiction with CAPM, if rationality exists, then CAPM is sound until you take into account efficient markets, however if rationality is not a permanenet state in the market, then CAPM cannot exist, so whether rationality exists or not, CAPM has some huge problems. But I wasn't clear on that in my exact point before.

I also do not like random walk. I almost exclusively use technical indicators and price action in my trading, so I go against everything that random walk states. However, it does sound good from a theoretical standpoint which I understand as this:

Prices exist as a consensus agreement of the value of the company on which the security is being traded, this consensus value is reached through independant analysis of all available news which could affect the value of the company by all of the participants of the security in the market. As a new piece of news emerges this consensus value with either change or will not based upon the consensus interpretation of the news, efficient market hypothesis tells us this should happen immediately. As it is impossible to know what the next piece of news will be as by definition it has not happened yet, it is impossible to know what effect the next piece of news will have on the price, it is also impossible to know when the event will happen. So prices should be random as each piece of news will not be known in advance.

The part about price charts having a similar appearance to a chart of random numbers merely strengthens the argument, but does not prove it.

I do not like the random walk theory because it completely ignores the irrationality in the market, and the fact that some people (such as me) will trade on price and not news. If you include the fact that some (actually most) people trade on price then ides like support and resistance become valid.

sorry for going on a bit but once I get started........
 
I too am primarily a technical trader, so I also have a fundamental bone to pick with most classical financial market theory as I was taught it in school. Trying to be objective, though:

One of the break-downs in EMT and Random Walk is that is effectively assumes that the market does not move between news events. Essentially, you have periods of flat interspersed with jumps as adjustments are made to new information. Great in theory, but we know that markets move even in the absense of new information - at least broadly available information anyway. There's a "Jump Diffusion Model" out there that I always thought did a better job of explain market movement. I don't recall the exact wording of it at the moment.

I recently read Mind Over Markets to give myself a refresher on the Market Profile method of analysis. The book brings up something that we as traders sometimes forget - that markets move to facilitate trade. A market will not long stay in a position where buyers and sellers cannot be matched up. That can be viewed as suggesting the markets find the level of shared relative value - meaning buyers and sellers come to mutual agreement as to the value of the instrument in question. The problem is people have different determinents of value, so most of the time there is an imbalance. At any given price it is rare that you have equal number of folks willing to buy and sell.

Some of the most interesting current thinking on financial and economic theory comes from the area of Complexity. I've only begun to scratch the surface there, but it's made the most sense to me of anything I've ever been exposed to from an academic/theoretical basis.
 
Correct me if I'm out of line here but don't you need to have 1500 hours of actual experience for the CAPM and 4500 hours of experience for the PMP title? To say that you have no experience flies in the face of the program itself. Due to my being outside the field for so many years now, I only qualify for the 1500 hour CAPM title, I'm no longer able to qualify for thr 4500 hour PMP title simply due to the time restriction of 6 or 8 years, I can't remember the exact time-line
 
Top