Press "Enter" to skip to content

Behavioral Finance Application

Behavioral Finance Application
Behavioral Finance Application

pBehavioral Finance Applicationp pso the theory of behavioral finance sounds great but in the end we need to figure out how do we actually implement this stuff well so weve developed a fivestep process that is designed to systematically exploit behavioral bias at each juncture along the way and this is not to say that this is the only way one can do this and we are going to try to limit the detail to some extent so this doesnt become a 20 hour lecture but its to give us an idea of how one might actually implement this in practice now quantitative value effectively buys cheap high quality stocks the Wall Street hates thats the in state of what were trying to achieve here but its not because we just decide at the beginning that thats a good idea its because it was built upon through a variety of steps to try to exploit bias problems so it just turns out that these are the stocks that suffer from the most behavioral bias and tend to have the largest limits arbitrage as we discussed in the theory when we have that combination of bias as well as limited arbitrage we find mispricing opportunities in other words quantitative value is really a dream for those that actually believe in this behavioral finance stuff which of course we do now if youre one that believes that the market is perfectly efficient then what we describe here represents simply a story but not an empirically robust phenomenon so the biases we seek to exploit with quantitative value in particular are the following one lottery bias and this is a bias where people tend to overvalue lottery like payoffs now representative bias is all about naive extrapolation of short term trends and well talk about how that affects investor behavior and stock prices and then the next one is limited attention and limited attention simply refers to the fact that our minds have a limited ability to focus on all things all the time finally were left with availability or what you could call CNC bias and this is this bias that we had previously discussed which simply means that humans tend to overweight the probability events that are more available in their mind so if they see the same thing over and over again or sellside analysts or CNBC commentators or mentioning something that piece of information that is very available and easy to recall for us tends to get over waited in a probability sense alright so step one lets first look at lottery bias so heres a table from the maxing out paper which shows very clearly I think what happens to stocks that exhibit lottery like behavior and what the authors do is they sort stocks each month based on the highest daily return of stocks for that prior month those stocks with large one day returns are considered more akin to lotteries whereas those stocks that dont have this sort of price action or consider less like lotteries and what the what happens next is the authors identify these lottery characteristics and they sort the portfolios based on this and they find that investors on average over value the lottery type stocks which is represented here by the Himax portfolios and the associated alpha which controls for firm specific characteristics and a lesson from this paper I think is that we want to avoid zero or one type investments because other investors may over pay for these situations plus investments like this can cause a serious capital preservation problem now were going to talk in more detail in a later presentation about exactly how we think about avoiding lottery stocks but the lesson learned for the time being is that lottery stocks in general or something to be avoided so moving on to representative bias so the original paper on this subject comes from lakhani shock Slifer and vision a 94 which your journal finance paper which is commonly referred to as lsv and the authors wrote this paper to counter the conjecture that the reason value stocks earn higher returns over time relative to growth stocks is due to the risk inherent in growth stocks lsv made the claim that the value spread is not due to increased risk but due to behavioral bias and in particular lsv focus on the idea the investors suffer from representative bias and extrapolate good news about growth stocks into the far future and bad news about value stocks into the far future but because of representative shortterm patterns that cause investors to extrapolate way out and time different growth rates investors along that path fail to appreciate that these growth rates in the future tend to be highly mean reverting if we have a value stock and it starts having a poor run we extrapolate it down to negative infinity if we have a gross stock that has a good run we extrapolate that great growth out to positive infinity and of course systematically what happens is value stocks with poor growth rates tend to mean revert back to average growth stocks with great growth rates tenda mean revert back to the overall market ever but because the market fails to appreciate this fact value stocks systematically surprise the market so other people dug in the weeds of this lsv conjecture and confirm that the value anomaly is likely due at least in part to behavioral bias and not a hundred percent due to a riskbased phenomenon now the bottom chart shows the relationship between pass earnings growth and future earnings growth so as we see here these glamour or growth stocks in the lowest portfolio rank based on book the market have really high pass earnings growth but their future earning growth mean revert so it starts to weigh down meanwhile the cheapest decile of stocks the value stocks have done terribly recently actually have negative earnings growth but their future earnings growth mean revert back and this is a general pattern across markets where growth is a mean reverting phenomenon likely due to microeconomics story involved there where if a firm is making a whole bunch of money a lot of entrants will come into that marketplace driving the returns on the capital down similarly in markets where the world is falling apart a lot of players leave that space which allows the current people in place to earn and enjoy higher returns on their capital and what happens here and what the Shaolin Sloane papers show is that its the sellside security and analysts that fail to appreciate this meaner version and because the markets often follow the lead of this outside analysts value stocks continually surprised the market on average when they come out with the earnings estimates relative to the growth stocks which continually surprise to the downside when they come out with their growth estimate overall this is represented by the representative bias phenomenon heres our own study we look here at simple tercel sorts which means we just sort the universe based on EBITDA tol enterprise value for mid and large cap firms from 1982 2013 and what we see is that there is a clear relationship between price paid in future performance now one could argue that value stocks are much riskier than growth stocks and therefore warrant the nearly four percent premium for low price stocks versus high price stocks however one could also argue that much of this premium earned by these value stocks is really more associated with represented bias in the marketplace investor irrationality so next well look at limited attention so Pete Rose can so suggest that investors fail to appreciate how company fundamentals relate to returns or when I mean company fundamentals were not talking about future expected growth rates now were talking about balance sheet items income statement items things that we can look at that have happened in the past and it seems as though investors bucket firms with similar price characteristics into the same bin but fail to recognize differences in their underlying fundamentals or what we call quality for example you might have IBM at a 10 x price to earnings ratio and you may also have made often company at a 10 x price to earnings ratio the example here is illustrate that madoff would be in theory a worse off company than IBM least fundamentally even though both are considered value stocks for illustrative purposes consider that IBM has stronger balance sheet and income statement items and may have actually has some money in the bank but not that much well it turns out that this what i would consider somewhat obvious element of stock prices is overlooked by market participants who are so focused on the news about future growth rates and the opinions of sellside analysts that they forget to focus on fundamentals so what piotrowski show in particular in their study is that the value growth strategy which is simply the spread between value stocks and growth stocks can be split based on fundamentals so if we do a strategy that is long value stocks that have great fundamentals and short grow stocks that have terrible fundamentals you get the returns associated with the straight line if we do the opposite we do a strategy that is long low quality value and short high quality growth we get the dotted line what you notice is that the straight line does really well on average the dotted lines basically right around zero and then the black lines which is just the standard value gross strategy kind of somewhere in the middle and what the intent of this analysis is meant to show is that investors are failing to appreciate how fundamentals relate to the returns associated with value and growth strategies and the bottom line lesson here is that when looking within value stocks we should think about fundamentals again we do our own quick and dirty study of this and we say how does this phenomenon play out via simple stock sorts on price and quality so for price we use our EBITDA total enterprise value sort and we move stocks into three bins and then we further sort those price bins in two different quality bins based on our quantitative value algorithm which is beyond the scope of the current discussion but its essentially a computerized version of what Graham does in security analysis just looking at balance sheet income statement items to see if this company is is of higher or lower relative quality what you look at in the chart here is that there seems to be a general increase in returns associated with price so if you pay a high price you tender lower returns and if you pay a low price which are these green bars we also notice is there seems to be a return premium associated with quality were lower quality firms holding constant price tend to earn less expect to return than high quality firms this effect is not as strong as price but it definitely seems to be there this of course is completely counterintuitive since lower quality firms are presumably riskier and therefore should earn higher expected returns on average we see the opposite and we attribute this quality premium to eliminate attention bias specifically related to investors limb attention and focus on balance sheet strength and firm fundamentals so moving on now to the final piece of the puzzle that that we like to use and that is how do we exploit availability bias so before we dive in on the particulars of how we exploit availability bias lets just discuss some research has already been done by various academics now if you remember availability bias is this situation where things that are available or salient in our minds are considered highly probable even though they may not actually be any more likely than things that are less available in our minds so for example if we just endured an earthquake even though maybe unconditionally an earthquake probability is 1 percent were now going to think that earthquakes are going to happen every single year the same thing happens in investing so if we look at this study by Fang and Perez which is 2009 journal finance paper this study the authors look at the performance of stocks with high media coverage relative to no media coverage after controlling for common characteristics that we already know explain returns so controlling for market beta size of the firm exposure to value momentum and so forth how does media coverage affect returns and the result is fairly striking high coverage firms down here slightly underperform whereas no coverage firms tend to earn excess returns and the argument behind this phenomenon is its an availability bias problem and the logic is fairly straightforward is it more likely that a stock that is on cbc every day and being talked about by all the Wall Street analysts is overvalued relative to a stock that no one has ever heard of because of availability the argument is that they stock that is on CNBC each day will have information that is highly available to market participants and therefore they increases the prod they may form incorrect probability assessments when they produce analysis about the CNBC stock relative to the stock that no ones ever heard of how would we operationalize this idea given data constraints and some of our own inhouse analysis on ways to exploit vail ability biased one of the things we like to look at is sellside opinion sellside analysts are often the talk of Wall Street and their information is highly available on the street everyones reading the report everyones talking about it on Wall Street so this seems like a good opportunity to perhaps exploit availability bias by focusing on sellside so we hypothesized that when Wall Street opinion was concentrated in one direction versus more dispersed wed be able to capture an availability premium the general concept is that when all the sellside analysts are bearish on the security and they are all citing the same reasons for a downfall these stocks might suffer from availability bias more than stocks where the sellside opinion is more dispersed and theres not a single story coming out that can influence investors one way or the other so for illustration what we do is we break out the portfolio that is the hot cheap high quality portfolio and we break it out into the high optimism and low optimism buckets and the results are pretty clear the cheap high quality stocks that Wall Street hates and presumably the stocks that are suffering the most from availability bias tend to do very well in the future whereas those cheap high quality stocks that Wall Street cant decide on where presumably people are suffering less from availability bias tend to perform just a little bit better than the average market so it does appear that availability bias contributes to this excess return associated with the low optimism or the stocks of Wall Street hates so to recap this was an explanation of how one might apply behavioral finance in financial markets and our job is to try to exploit as much bias as we possibly can in a systematic way and we have sidestepped admittedly the discussion of limits arbitrage in this presentation mainly due to time constraints and a desire to save some of our internal research for the benefit of our own investors nevertheless as a general rule the Securities generated by the process just described are by construction the stocks that many institutional investors wrinkle their nose at and discuss they think we are insane for wanting to own the cheap high quality stocks at Wall Street hates we think we are simply practicing sound behavioral finance by building systems that beat behavioral biasp

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *