|
On June 08 2009 19:34 Automatically wrote:
actually, both CAPM and APT are fundamentally flawed. In fact , its the inverse, CAPM does hold empirically, to some extent if we are being lenient, but it remains incredibly flawed theoretically speaking. Buffett made a strong point about this issue. His argument is so simple, powerful, intuitive and elegant, that i'll let you discover it yourself. Basically, historical volatility should never be taken as a proxy for riskiness. I understand that its a reflexive intuition to use it as such, but its just plain wrong to explain future risk with such a strong emphasis on historical prices. You can come up with any fancy model you want, all the APT's of the world, but statistics are only statistics. i.e. a mathematical representation of the past . no one can predict the future with certainty. Business is business. That world is affected by way too many factors, most of which are not knowable or comprehensible to even the most brilliant human minds. if you're too confident in your predictive powers, you end up broke like LTCM.
Also, 100% equity > all , especially for young people like us. and even more especially because of today's prices.
Explain to me how the theory is fundamentally flawed. There's no mistake in the mathematics. All I've stated is that the empirical fit can be dubious.
CAPM relies on the Market portfolio containing all assets in existence, it requires identical investors such that everyone faces the same CML, it requires equilibrium in the market such that the Market portfolio already contains all assets with no excess demand for each, it extends from Mean Variance analysis which implies a requirement of normality of returns or quadratic preferences. APT on the other hand generalizes CAPM to a multi-factor model (since CAPM implies the only factor you need to consider is the Market) but because the Market porfolio is unobservable in the real world, APT's generalization allows any other factor to be incorporated while making no strong assumption on investor preferences, makes no requirement on pricing all assets in existence etc.
Have you seen the empirical literature? Let's take the standard Sharpe's One Factor model for the CAPM - the standard regression of y = a + xb + e where y denotes the excess portfolio return, b is the asset's beta, a is the intercept and e is the residual. CAPM implies a = 0, the coefficient of b, x, should be the excess market return, only the asset beta should be determining asset return and the model should be linear. The vast CAPM literature says a is greater than 0, x is less than the excess market return and while beta is a good explanatory variable other factors should be taken into account. This is most likely due to proxying the Market portfolio with a market index. APT's empirical fit is better, like Fama-French's 3 factor model. A paper I like to cite is Chen (1983) where he shows APT can explain much of the residual variance in CAPM regressions but not the other way round. So what's your empirical evidence that CAPM holds while APT doesn't?
Volatility needs to be measured somehow. As I said before, use of past volatility is because volatility is correlated - this is an empirical fact. When you forecast you don't make a straight line forecast, you take into account your uncertainty by forecasting within a range of outcomes since you have to make an assumption/estimation about the distribution of returns. No one says volatility forecasting leads to certain results.
Did you read and understand why I said 100% equity may not be desirable? Sure, you may get a nice return in the long run, but it comes down to the classic problem of whether you can stay solvent long enough to do so. Especially during times of high volatility. For example, if you've got everything in an asset and it does badly in the first time period then your consumption level is going to take a huge hit in that period, regardless of how well you think it's going to do later on. Since most people don't want to be consuming very little in any given time period, it implies portfolio diversification is a good idea.
Edit: I see I have something else I need to respond to while I was typing this out. I'll get round to it later, but for now I'll say volatility risk for a portfolio is taken to be the variance of the portfolio because for a well diversified portfolio there should be no idiosyncratic risk.
|
Calgary25977 Posts
|
I know this wasn't directed towards me but I can't help but to throw a few jabs.
On June 09 2009 01:57 Milton Friedman wrote:
Explain to me how [CAPM] is fundamentally flawed. There's no mistake in the mathematics. All I've stated is that the empirical fit can be dubious.
--There's no mistake in the mathematics but the assumptions which the math derives from are false.
Volatility needs to be measured somehow. As I said before, use of past volatility is because volatility is correlated - this is an empirical fact.
--Sure, volatility can be measured; nobody is arguing that beta is not a good measure of volatility. But volatility IS NOT risk, whereas the standard interpretation of the models suggests that it IS.
Did you read and understand why I said 100% equity may not be desirable? Sure, you may get a nice return in the long run, but it comes down to the classic problem of whether you can stay solvent long enough to do so. Especially during times of high volatility. For example, if you've got everything in an asset and it does badly in the first time period then your consumption level is going to take a huge hit in that period, regardless of how well you think it's going to do later on. Since most people don't want to be consuming very little in any given time period, it implies portfolio diversification is a good idea.
--Again, a rational argument based on a false premise. Most young people do not rely on their investments for consumption income, they rely on them for capital gains. Even if we step away from the assumptions about the investing rationale of a specific demographic, the simple fact that the intelligent investor CAN ignore short-term unrealized gains/losses is enough to disprove the necessity of equating risk with volatility. My solvency is not at risk if I don't employ leverage (because I'm not deluded into thinking my models are perfectly correct).
for now I'll say volatility risk for a portfolio is taken to be the variance of the portfolio because for a well diversified portfolio there should be no idiosyncratic risk.
[Emphasis added]. Sure, but there's market-wide systematic risks which screw up the predictions of the models by turning historically uncorrelated assets into correlated assets.
I'm not responding to the rest of your post because it misses the point. Nobody doubts the mathematics behind the models, what we're saying is that the assumptions that drive those mathematics are false. I can state that "apples are blue" and then follow it consistently by saying that "apples are the color of the sky"; which would be true, if apples really were blue, but they're not! Similarly, these models can draw impressive mathematical predictions about the movement of real assets IF volatility really does equal risk, but it doesn't!
|
|
|
Kewlsunman, how do you initially pick a stock to study? Do you use a screener and enter the different criteria you're looking for? I must imagine all these analyses take time, so there must be a good way to have a grouping of stocks that pique your interest right off the bat.
|
On June 09 2009 11:51 pokeyAA wrote:
Kewlsunman, how do you initially pick a stock to study? Do you use a screener and enter the different criteria you're looking for? I must imagine all these analyses take time, so there must be a good way to have a grouping of stocks that pique your interest right off the bat.
Yep, I might've mentioned this somewhere in the guide but I'll go over my filtering process:
1) I start off with a stock screener; I like Google's but maybe I'm just used to it. It has most of the things I want to look for but there are some other ratios I wish it'd include (cash flow stuff mostly). One set of criteria I might look at would be market cap. < $500 mil, P/E <10, ROE >8%, Debt/Equity <20%, 5 yr Net Income growth > 8%. This would come up with a list of decently run "growth" stocks which might become something interesting in the next 3 to 5 years. Or I might just look at stocks that are trading at 52 week lows, and just try to find a stock that is severely temporarily mispriced.
My way isn't necessarily the "right" way though. I'm at a stage where I'm still building up my business database, so running screens is still helpful. There are other professionals I know who simply work down the list of all US publicly traded companies, alphabetically or by region. There's no way to get around the fact that if you really want to get good at this game, you'll need to know the entire field. Doing screens is sort of "cheating" in that respect; the real good stuff doesn't get picked up so easily. Because what you have to keep in the back of your mind is that if you can do a screen, so can other people, so it's very rare that just through running a screen will you come across something that's under-followed and subsequently undervalued.
2) Once I have a list of 30 or so from my stock screen, I'll go through the most recent financials and ratios of each one. Most finance websites have that information very handy so it's easy to just quickly browse through. I'm mostly looking to see if there's any severe discrepancies with what I'm looking for. Usually I can screen out anywhere from 1 to 5 just from this quick lookthrough.
3) With the remaining 25, say, I'll go and read their most recent annual reports, one by one. I want to look to see if I can understand the business, if I can understand the numbers, and again, if there's anything that's just out there that doesn't make any sense. I can narrow down the list to about 5 or 10 here. Lately, for instance, I've been coming across a lot of natural gas and oil trusts which basically get a portion of the proceeds from the sale of natural resources on the land they own. To invest in those, I'd have to be able to confirm the accuracy of their oil production predictions, I'd have to be confident in management's ability to find new fields, etc.--things that I don't think I'd feel comfortable doing.
4) With the final 5 or 10, I have a giant spreadsheet that I'll use to plot all the historical data I can find. Again, by this stage, the company has passed the initial qualitative screen; I think it's a company I can understand with financials that seem to make sense and are consistent with how I imagine the company should be doing. At this point, I want to go through their history, look at their track record, their past financial results and start painting a picture of how this company has been run, previously, and how it might look to be run in the future. I will generally find a lot of curious things in this stage (large cash outflows that I need to make sure are not consistent recurrences, asset fluctuations that I need to make sure are adequately explained, etc.) and I'll have to read through most of their past annual reports to paint a timeline of the company's history up until this point.
5) By the time I've gone through all that, I'll usually be down to 0, 1, or 2 companies which I still find interesting. At this point, if I haven't already, I'll find some industry periodicals where I can get a better feel for how the company does within its field. While I'm not overly focused on macro issues, it may be a good time to just check to see that the entire industry isn't going down the drain anytime soon, stuff like that. And only at this point will I look to see if there are any other opinions on the company; other research reports, analyst opinions, etc. Really I just want to make sure that I haven't missed anything big in my research; I'm not really interested in other people's conclusions, I just want to see if we've come across the same facts (I trust in my own decisions). This is also where I might try to get in contact with the company's investor relations department, either by phone or email, if I have some important questions, and also, again, just to make sure that my facts seem right.
6) Only after doing this research and checking and rechecking my facts will I attempt to make an investment decision. I'll do a DCF valuation, an asset readjustment valuation, or whatever else I feel is appropriate in this case, just to see if the price is right. If I feel that the price has a nice margin of safety to intrinsic value, I'll buy. If not, I'll add it to my watchlist and then move on.
|
|
|
Man, I guess after experience this must come a lot faster haha. I'm just starting to learn things and I must admit, the number of metrics and etc to look for is a bit daunting. One quick clarification to make sure I'm getting this, for your analysis on EPV, am I correct in understanding that youre taking past earnings percentages and forecasting them into the future? I have this random tidbit ringing in my head where "past performance does not guarantee future results", so the margin of safety is used to counteract the dangers of forecasting?
Thanks a lot by your contribution so far, its been really helpful.
|
Excellent job and very well written!
I'd just like to make two quick minor points though. First is Eddie Lampert is a massive luckbox and doesn't deserve to be on your important people list when a true genius such as James Simons is nowhere to be seen.
Second point that I'd like to make is to encourage people to take an interest in personal finance which was not mentioned anywhere here since it is a seperate topic altogether. However, doing a good job at managing your personal finances applies to everyone and is a relatively simple task and yet, many highly intelligent people fail at this endeavor. Managing your spending properly, and making wise financial decisions such as limiting the amount of debt on high interest credit cards, avoiding overdraft fees, and maxing out on tax shelters such as 401K and IRAs just to name a few are not as sexy as mastering the stock market, but it is the foundation upon which your entire financial life will be built upon.
|
On June 09 2009 14:25 SaveMySoul wrote:
P/E less than 10? lol, i love that. but you're not gonna find a growth stock with that kind of ratio.
Not your traditional growth stock that everyone on Wall Street is already talking about, no. But an under-covered company with 3-4 years of operating history with a strong catalyst? There's definitely a couple of those out there.
On June 09 2009 14:28 pokeyAA wrote:
Man, I guess after experience this must come a lot faster haha. I'm just starting to learn things and I must admit, the number of metrics and etc to look for is a bit daunting. One quick clarification to make sure I'm getting this, for your analysis on EPV, am I correct in understanding that youre taking past earnings percentages and forecasting them into the future? I have this random tidbit ringing in my head where "past performance does not guarantee future results", so the margin of safety is used to counteract the dangers of forecasting?
Thanks a lot by your contribution so far, its been really helpful.
I'm not quite sure what you mean by "earnings percentages" but the theory behind EPV is that you're trying to put a value to a consistent annual earnings level. So, if you figure that Company XYZ can earn $100,000 a year, divide by a cap. rate of 10% (say), and you have a rough EPV of $1,000,000. There are two variables here that you need to input: a consistent future earnings level, and a cap. rate. Using 10-15% as a cap. rate is fairly standard practice (and it's decently conservative), although I do encourage you to think a bit about why those might make sense.
As for projecting a consistent future earnings level, there are a million ways and different variables which might affect how you project it. Past earnings levels may affect your decision, so might certain qualitative catalysts which you think might have a strong impact on sales, changes in expenses, etc. If you have a relatively stable company that isn't changing a lot in the near future, then it's not unreasonable to take the average of their earnings for the past 5 years as your projected future earnings level.
I think it's great that you are keeping in mind that "past performance does not guarantee future results", but it's silly to imagine that they might not have some correlation. The important thing, though, is to figure out the fundamental factors that caused the past performance, and then see if those will hold in the future.
|
On June 09 2009 14:58 tenbagger wrote:
Excellent job and very well written!
I'd just like to make two quick minor points though. First is Eddie Lampert is a massive luckbox and doesn't deserve to be on your important people list when a true genius such as James Simons is nowhere to be seen.
Second point that I'd like to make is to encourage people to take an interest in personal finance which was not mentioned anywhere here since it is a seperate topic altogether. However, doing a good job at managing your personal finances applies to everyone and is a relatively simple task and yet, many highly intelligent people fail at this endeavor. Managing your spending properly, and making wise financial decisions such as limiting the amount of debt on high interest credit cards, avoiding overdraft fees, and maxing out on tax shelters such as 401K and IRAs just to name a few are not as sexy as mastering the stock market, but it is the foundation upon which your entire financial life will be built upon.
Great nickname :D
I actually do mention James Simons (or at least Renaissance, maybe not him by name) in another post in the thread, but not in the original guide. I'm not sure if I agree about Lampert being just a luckbox, but I certainly agree that he's not yet on quite the same level as some of the other luminaries there. Although, I am very interested to hear why you think he's so overrated!
Thanks for your post!
|
|
|
I don't mean to say that Lampert isn't a very intelligent and qualified investor. But I think he's highly overrated. He made a genius move with Kmart/Sears and made billions on it and the press was all over him hailing him as the next Warren Buffett. But that deal didn't turn out nearly as well as it seemed a few years ago and he made colossal mistakes over the past few years such as his huge build up of Citi stock.
What makes Buffett so amazing is the consistency of his returns. Anyone can get lucky by making a huge bet and winning. Few can create very robust models that outperform year after year. James Simons is the very definition of robust and his performance year in and year out, in good markets and bad is truly remarkable.
|
On June 09 2009 15:42 SaveMySoul wrote:Renaissance and simons and one of the very very few hedge funds that can boast such high returns based on a global macro or complex statistical models. Considering that they are over 6000 hedge funds in the US, and that some significant proportion of them are using a global macro strategy, then yes, its logical that a few of them will be in the spotlight due to a 30% + average returns purely based on the sheer numbers of the running horses. Lampert is truly one of the best. Probably Buffett's successor. One of the best investors of his generation. Your list is good and i like how its biased towards value investors (and it should be)  however, i do think you should mention Soros just because this man is so huge, even though i really hate successful speculators.
So out of the entire universe of hedge funds, Simons was #1 in 2006, #3 in 2007 and #1 again in 2008. That cannot be attribute to "purely based on the sheer numbers of the running horses".
And we're not talking about 30%+ returns here.
"Jim Simons continues to defy logic. In a year when most hedge funds lost money — and the few that were in the black generally managed less than $1 billion — Simons, the renowned founder of East Setauket, New York–based Renaissance Technologies Corp., generated a hard-to-fathom net 80 percent return at his 20-year-old flagship Medallion Fund. What makes this feat even more incredible is that Simons, one of the members of Alpha ’s inaugural Hedge Fund Hall of Fame (June 2008), charges a fat 5 percent management fee and 44 percent performance fee. To put it another way, Medallion — which has about $7 billion in assets — was up almost 160 percent before fees."
What was Lampert doing in 2008 and 2007? How about losing a billion+ in back to back years?
http://www.iimagazine.com/Alpha/Article.aspx?ArticleID=2165674
2 Edward Lampert
ESL Partners
↓ 2008: $1 billion
↓ 2007: $1.1 billion*
ESL’s concentrated portfolio was down 35 percent. Its largest position, Sears Holdings Corp., was down more than 60 percent.
|
A few interesting comments regarding Simons on a forum:
"Here is some interesting comparison data related to the 105,000% proported returns for Medallion over 20 years. If you purchased the spx at the low during the great depression, and then sold it at the high in 2007 your return would be 30,000% (over a period of 75 years) If you purchased MSFT at the all-time low and sold exactly at the all-time high, the return would be 60,000%. If you purchased GE at the all-time low and sold at the all-time high the return would be 6,000%. If you purchased XOM at the all-time low and sold exactly at the all-time high the return would be 3,000%. So basically Medallion is the greatest wealth creation vehicle of all time without having a down year in two decades or a down quater in one decade. Why are there 200 books about Warren Buffett and zero about Jim Simons?"
and the other side:
"In the post Madoff era, It would seem that Renaissance's return on Medallion (opaque) of 80% while RIEF (transparent) and RIFF (transparent) lose around 20% and 12% repectively leave open the possibility that Medallion's legendary returns were fiction all along. Where is the SEC and the NFA on this one? Medallion's supposed cumulative return since 1989 is 105,051% which is 20 times better than Tudor, 35 times better than Warren Buffett, and 20 times better than Soros. I challenge Renaissance to have the returns audited and make it public. Otherwise don't belive the hype. We must assume they are just an averge hedge fund that is cooking the books untill proven otherwise."
The disparity of performance between medallion, which is almost entirely owned by Simons and other employees of his firm, and the other funds which constitute institutional investor money is definitely curious. The Madoff scandal showed how we should always be skeptical of results too good to be true. While I acknowledge a definite possibility that something is not quite right here, IF Simon's reported results are indeed accurate, there is no doubt that Simons is GOD and the mere mention of Lampert in the same sentence is blaspehmy.
|
|
|
Just the sheer size of this guide is enough to get a beta key.
|
Wow, I definitely have to read all of this later, looks very interesting. Thanks for your hard work kewlsunman
|
On June 09 2009 16:30 SaveMySoul wrote:
im not denying Simons talent, im just saying how his methods can not be imitated by many others. Like i said, very very few people are successful by trading like he's doing, and in my humble opinion, its because that method is inferior to Buffett's value investing, which has produced more billionaire investors than any other.
Also, comparing him to Buffett is just dumb. Simons, or anyone else, can have a huge year, 160%. Thing is, Buffett's long-term record destroys everyone. He had tons of back to back 60% years, and many times he had about four 50% years in a period of 5 years. Also, Buffett suffers from what he calls a huge anchor to performance. The Dude is managing a 150 billion company in terms of market cap. (200 before the crisis) That has nothing to do with Renaissance's 7 billion. And therefore all his best ideas for Berkshire can only affect a small portion of his overall return. with a 150 billion cap, your return is greatly affected by your 15th to 35th best investment ideas which you are not forced to use when you're managing peanuts. Buffett said if he was managing peanuts, he'd average 3 digits returns every year.
That's probably why there are more books on Buffett than Simons or anyone else. But yes, Simons is indeed very smart.
I'm a huge Buffett fan and I read his annual report every year. The comment that compared Buffett to Simons was something I copy pasted from a different forum and I completely understand that their strategies are completely different. Finally, I recognize that Simons results are not transparent and there are question marks surrounding it.
With that said, let's go under the assumption that Simons reported results are indeed accurate. You talk about it like he had one good year. That is what Eddie Lampert has had where he makes a billion one year and loses a billion the next year. James Simons over the past 20 years destroys everyone and that includes Buffett in terms of ROI. When I say that, I am not trying to diminish Buffett's amazing achievements for he has many more years of operating a much larger portfolio in a publicly traded company which is arguably the more impressive overall feat. My original comparision was not between Buffett and Simons but between Lampert and Simons and my praise of Simons and his clear superiority over Lampert by no means diminishes any of Buffett's accomplishments.
But your comments like "one good year" or "average returns purely based on the sheer numbers of the running horses" about Simons performance is bordering on ridiculous. It's one thing to call him a fraud and say that he's cooking his books. It's another thing to say his reported performace of a 105,051% return since 1989, and no down year in 20 years and no down quarter in 10 years is anything short of extraordinary.
|
|
|
|
|
|
|
|
EPT
