Learning, Sharing, and Teaching > Investor Alley
Can't beat an Index
brooklynguy:
--- Quote from: skyrefuge on April 22, 2015, 11:06:56 AM ---I still don't get this. Essentially you seem to be saying that you don't trust any research that uses statistical analysis? You require data from every single unit of the entire population in order to not cry "survivorship bias"?
--- End quote ---
Yeah. The concept of survivorship bias, important as it may be for the prudent investor to be cognizant of, seems to be getting a little overhyped around here. I think its jump the shark moment was the creation of two duplicate threads dedicated to the topic in the span of two weeks.
First, survivorship bias is only one of the many species of selection bias, so the inordinate focus on it itself represents a bias in the selection of selection biases on which to focus!
Second, as skyrefuge pointed out, survivorship bias is merely a type of sampling error that can occur when doing statistical analysis. In its absence, it can't be used as a sword against any observational study (or, worse yet, any experiment) that reaches the conclusion that a given cause produced a given result.
--- Quote ---at least around here, buy-and-hold "value" investing is really the only strategy in which there remains any reasonable debate. There aren't many proponents of day-trading or technical-analysis left here, so it wouldn't be very interesting to study those strategies.
--- End quote ---
Perhaps you haven't noticed the Dual Momentum thread that's been raging on the front page of Investor Alley for the past two weeks?
Dodge:
--- Quote from: skyrefuge on April 22, 2015, 11:06:56 AM ---
--- Quote from: Dodge on April 21, 2015, 09:35:03 PM ---I gave my reasoning. Is it your assertion that the study adequately controls for Survivorship Bias, by looking at a single private investing group, which invests in a single type of strategy, over a short time period in which that single strategy type beat the market overall? If so, you don't understand the definition of Survivorship Bias.
--- End quote ---
For there to be "survivorship bias" in selecting this group, there would have to be other such groups who were analyzed but not published (due to not delivering the "right" conclusion). Or those other groups would have had to have died out, which seems unlikely, since there really isn't any pressure from performance requirements that would cause poorly-performing groups to be closed more frequently than better-performing ones.
--- End quote ---
I see this as further evidence of how elusive Survivorship Bias can be.
------------------------------------------
It is easy to fall victim to the Survivorship Bias. After any process that leaves behind survivors, the non-survivors are often destroyed or muted or removed from your view. If failures becomes invisible, then naturally you will pay more attention to successes. Not only do you fail to recognize that what is missing might have held important information, you fail to recognize that there is missing information at all.
http://youarenotsosmart.com/2013/05/23/survivorship-bias/
------------------------------------------
In explaining why they choose this private investing group in particular, they say:
"The site has been heralded in many business publications as a top-quality resource for those who can attain membership (e.g. Financial Times, Barron’s, Business Week, and Forbes)."
I suspect sites who gave poor-performing advice, didn't end up in the Financial Times, Barron's, Business Week, or Forbes. If so, they fell victim to Survivorship Bias.
brooklynguy:
--- Quote from: Dodge on April 23, 2015, 11:43:54 PM ---I see this as further evidence of how elusive Survivorship Bias can be.
--- End quote ---
Dodge, you seem to have become so enamored in the idea of the pervasiveness and invisibility of survivorship bias that you are starting to see it where it doesn't exist.
I agree with you that there may have been survivorship bias in the selection of this group (and have less confidence than skyrefuge that there was no survivorship bias in that respect) -- maybe there were other otherwise-equivalent investing clubs which died out because they sucked at stock picking? Or maybe there was another form of selection bias at work in the selection of this sample-group.
But I agree with skyrefuge in that I see no reason to believe there was survivorship bias in the selection of the ideas produced by this group.
Essentially, the methodology these researchers used was as follows: they looked at stock-picking ideas produced and self-ranked by a group of investors (which admittedly may not be representative of the investing population as a whole), and examined whether the high-ranked ideas outperformed the low-ranked ideas, and found that they did. (Thanks to skyrefuge's sleuthing, we now know that there was a subsequent study that gives us reason to suspect that there may have been other flaws in the initial study's methodology, and I look forward to seeing whether JoJoK has any light to shed on that issue.)
To extend skyrefuge's vaccine analogy, this study was analogous to examining a subset of the population, some members of whom were given the vaccine, and some members of whom were not, and finding that a statistically significantly higher percentage of the non-vaccinated members contracted the disease. Is it your view that "survivorship bias" means we cannot trust those results to show that the vaccine has the effect of reducing a group-member's chances of contracting the disease? It's true that if there was selection bias in the choosing of the sample group, the results cannot necessarily be extrapolated to the population as a whole (e.g., if everyone in the sample group happened to be a Cherokee Indian, then perhaps the vaccine would not have the same effect on the wider, genetically-distinguishable, population). But it is reasonable to conclude that within the sample group, the vaccine either caused disease-resistance or was correlated with some other factor that caused disease-resistance.
Do you have any reason to believe there was survivorship bias in the selection of the ideas produced by this group? It seems to me that the entire universe of stock-picking-ideas produced by the group was examined, and the researchers found that the high-ranked ideas outperformed the low-ranked ideas. If there were no other flaws in the methodology, it would be reasonable to conclude that the high-ranked stock-picking ideas outperformed the low-ranked stock-picking-ideas because the stock-picking-idea-rankers were actually good at picking stocks in a better than random way.
Dodge:
--- Quote from: brooklynguy on April 24, 2015, 09:14:41 AM ---
--- Quote from: Dodge on April 23, 2015, 11:43:54 PM ---I see this as further evidence of how elusive Survivorship Bias can be.
--- End quote ---
Dodge, you seem to have become so enamored in the idea of the pervasiveness and invisibility of survivorship bias that you are starting to see it where it doesn't exist.
I agree with you that there may have been survivorship bias in the selection of this group (and have less confidence than skyrefuge that there was no survivorship bias in that respect) -- maybe there were other otherwise-equivalent investing clubs which died out because they sucked at stock picking? Or maybe there was another form of selection bias at work in the selection of this sample-group.
But I agree with skyrefuge in that I see no reason to believe there was survivorship bias in the selection of the ideas produced by this group.
Essentially, the methodology these researchers used was as follows: they looked at stock-picking ideas produced and self-ranked by a group of investors (which admittedly may not be representative of the investing population as a whole), and examined whether the high-ranked ideas outperformed the low-ranked ideas, and found that they did. (Thanks to skyrefuge's sleuthing, we now know that there was a subsequent study that gives us reason to suspect that there may have been other flaws in the initial study's methodology, and I look forward to seeing whether JoJoK has any light to shed on that issue.)
To extend skyrefuge's vaccine analogy, this study was analogous to examining a subset of the population, some members of whom were given the vaccine, and some members of whom were not, and finding that a statistically significantly higher percentage of the non-vaccinated members contracted the disease. Is it your view that "survivorship bias" means we cannot trust those results to show that the vaccine has the effect of reducing a group-member's chances of contracting the disease? It's true that if there was selection bias in the choosing of the sample group, the results cannot necessarily be extrapolated to the population as a whole (e.g., if everyone in the sample group happened to be a Cherokee Indian, then perhaps the vaccine would not have the same effect on the wider, genetically-distinguishable, population). But it is reasonable to conclude that within the sample group, the vaccine either caused disease-resistance or was correlated with some other factor that caused disease-resistance.
Do you have any reason to believe there was survivorship bias in the selection of the ideas produced by this group? It seems to me that the entire universe of stock-picking-ideas produced by the group was examined, and the researchers found that the high-ranked ideas outperformed the low-ranked ideas. If there were no other flaws in the methodology, it would be reasonable to conclude that the high-ranked stock-picking ideas outperformed the low-ranked stock-picking-ideas because the stock-picking-idea-rankers were actually good at picking stocks in a better than random way.
--- End quote ---
It looks like we're on the same page brooklynguy. There were two decisions here. The decison to choose this particular investing group, then the decision of how to analyze the data. I only see Survivorship Bias in the former.
CCCA:
--- Quote from: frugalnacho on April 21, 2015, 12:24:51 PM ---
--- Quote from: Chuck on April 21, 2015, 11:59:49 AM ---
--- Quote from: Dodge on April 15, 2015, 10:41:10 AM ---
No timing needed? Tell that to the person who bought in AAPL 1980 and underperformed the market for 26 years.
--- End quote ---
Brah. Check your own graph.
Someone who bought AAPL in 1980 is currently outperforming the market by nearly AN ORDER OF MAGNATUDE. See how the indicators on the Y Axis get closer together towards the top? That's because AAPL's outperformance of the market is too large to show to scale. It won't fit on your screen. I don't feel you've supported your argument.
The fact is that there are companies that outperform the market over the long term. IBM is one. Coca-Cola is one. Altria Group is one. Johnson and Johnson is one.
It looks increasingly likely that Apple will be one as well. It sure as hell is over the past 35 years.
Index funds are by far the safer bet. You are guaranteed not to pick a loser. But don't pretend that picking winners is impossible.
--- End quote ---
I checked his graph and it looks like apple did in fact underperform the market from 1980 until 2006 - 26 years, just as he claimed. It's only since 2006 that the stock has had explosive growth and beat the market. I think dodge's point is that in 2015 it's easy to look at apple and say you should have invested in it before the explosive growth, but who knew that back in 1980, or even 2006? You could have just as easily invested in Turd Inc. in 1980 and had your entire portfolio disappear along with the company.
Someone will be on an internet forum in 20 years making similar claims: "It's so easy, all you had to do was invest in Widgets Inc. in 2015 and you'd be a gajillionaire today!"
So what is the next apple that is going to make us all filthy rich?
--- End quote ---
Not necessarily filthy rich, but a reasonable person could have concluded that after the introduction of the iPhone in 2007, that maybe Apple would have a big hit on it's hands. And if someone had invested in AAPL after the Steve Jobs announcement, the stock price has appreciated by >800%, whereas VTI (total stock market) is around 54%. Even if you waited a couple of years (beginning of 2010) just to see that the iPhones were, in fact, selling, you would have seen a 350% appreciation vs the market (96%).
I'm lucky to have invested in Apple during its earlier and more recent growth. But I also know that lots of people don't "get" Apple and its stock. There's always been the narrative that since Samsung/Dell (or whoever) can put the same or even "better" parts in their computer/MP3 player/phone and sell it for cheaper, that Apple is always about to be beaten and fail. But this narrative, perpetuated by the media and analysts, always struck me, and many others, as being incorrect. So some of that "luck" has been the result of learning as much as I can about these companies and markets and also faith that over time this narrative will fade.
I will say that it never occurred to me that Apple would one day be the biggest company on the planet.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version