How Bloggers Beat Wall Street | Hacker News Reader

How Bloggers Beat Wall Street

blog.vuru.co

26 pointsvirincognito24 comments

Most of the top predictions for Apple's "surprise blow-out quarter" were from the finance section of a message board for Apple enthusiasts. This is like saying the New York Jets' message board is smarter than Vegas sportsbooks because they predicted an upset win over the Patriots.

Hi jhamburger, thanks for the comment!

I'm the author of the article and want to present a different perspective:

I don't think you're being completely fair to these guys. Sure, they're Apple enthusiasts but they're also digging significantly into the data to back up their analysis. For example, check out one users' analysis on Apple's Cost and Cost Ratios to Revenue: http://www.postsateventide.com/2010/11/apple-quarterly-compa... (Found here: http://www.macobserver.com/tmo/forums/viewthread/79667/).

OK so they're not guessing, but they're still biased. I'd like to see them correctly call a bad quarter before I'm convinced.

When was the last time Apple had a bad quarter? :-)

1997?

http://www.windowsitpro.com/article/news2/apple-posts-708-mi...

"The red ink just won't stop flowing from Apple Computer these days, as the company posted a tidy US$708 million loss this quarter. This is only one year after posting a $740 million loss in the same quarter of 1996. Apple said the loss contains a one-time $155 million charge for restructuring and a $375 million write-off to cover the acquisition of NeXT...

"We will return to profitability," said [CEO] Amelio, apparently to himself. Everyone else had stopped listening, except of course "MacWeek," which reported this story as "Company reports $1.6 billion in revenue." I guess that's one way to look at it"

(Fun fact: if you'd bought $10,000 worth of Apple shares in 1Q of 1997 they'd be worth eight hundred thousand dollars nowadays.)

Though I liked your comments in the article, I agree with jhamburger, this is a lot different than taking a random sampling of the popular finance bloggers and comparing their top picks vs. the top picks of analysts. You could still allow the bloggers to choose the companies they know best, but you would get more realistic picture of their investing prowess than the results here. That said, I very much agree with the 4 points you stated in the article.

The only people who will be surprised by this are people who do not understand probability.

Nobody expects analysts to be right every time, and nobody expects individuals to be wrong every time. You can't extrapolate a general principle from a single datum.

Now, if this were happening consistently -- amateurs were consistently more accurate than professionals -- then I would be surprised. I would draw the conclusion that the professionals had some other goal than accurate predictions, though, rather than that they were incompetent.

I have several problems with the CNN article this one is based off of.

1. Why take a screen shot of the spreadsheet?

2. Why are there only 11 amateurs? Amateurs greatly out number pro but in this table they're a minority. Seems like sample bias to me.

3. Where is the context and the history? The bloggers all predicted larger increases then the pros. Do bloggers always predict higher numbers? If so it's no surprise they did this time.

Surely bloggers do not only always outperform Wall Street analysts, but there are certainly scenarios, such as this one, where they do get it right. I agree that CNN's sample size could be statistically inaccurate, but the blogosphere is pretty difficult to measure.

I think the point of the article is that the number of individual investors is growing, and since there is such a plethora of information out there and easy channels to share opinions, the financial sphere is rapidly evolving online.

Surely bloggers do not only always outperform Wall Street analysts, but there are certainly scenarios, such as this one, where they do get it right.

s/bloggers/chickens/

Seriously, if I have a bunch of chickens pecking at labelled buttons to predict corporate profits and I only bother to report those cases where they happen to be right then my chickens are gonna come out looking pretty good as well.

I think you're missing the point. The argument isn't that bloggers as a whole perform better than professional analysts, it's that there are far more sources to get information about the market these days than just Wall Street analysts. Bloggers are one of them, and there are times that they are right.

Some blogs are about as reliable as pecking chickens, but some spend considerable time researching and developing solid insights. Generalizing and discounting every single one is a bit rash.

But unless there's someone whose advice is reliably better than the market, then merely knowing that some people are better some of the time isn't useful information.

There are a lot more bloggers than professional stock analysts. Therefore, doesn’t it follow that some bloggers would be the most accurate based on random chance alone?

Well, if in the top 10 there were only a couple bloggers, that would be a reasonable statement, but the top 9 were all bloggers. That doesn't sound like random chance to me.

Well, if there were a million bloggers posting random predictions and only one analyst, it would seem, intuitively, that the top nine would likely all be bloggers. It all depends on the number of bloggers that made predictions but aren't in your list. Is it zero? A trillion? Given that most bloggers tend to advertise their successes but not their failures, how can you get a definitive number?

Bloggers outnumber AAPL sell-side coverage by infinity to 48.

i think a t-test is in order

There might be another issue at play here: perhaps the investment firms represented by the professional analysts have incentive to under-predict a company's performance in their public-facing reports.

Yeah, my friend who is more knowledgeable than myself in equity trading said this a few days before apple reported earnings:

"The reason is so that on earnings day, AAPL can crush "consensus" earnings estimates and have their stock price increase. This is what Steve Jobs wants to happen, and the street analysts are happy to fall in line so that he'll continue to meet with them (or their clients) and/or do business with their firm if Apple ever needs investment banking advice

The next question you might ask is: doesn't this look bad for Apple if people are projecting worse earnings into the future? Don't stocks trade loosely on things like P/E ratios?

The answer is that if you look out a full year, the effect is actually the exact opposite, analysts tend to be way too optimistic (link<http://www.ritholtz.com/blog/2010/06/mckinsey-equity-analyst...; ).

You can see how this would work. You look at earnings estimates a year out and think: "man, this stock looks pretty good if earnings are going to grow X% over the next year". And then you look at the next quarter results and say "man they did better than expectations! This stock must be REALLY good. Maybe they will grow at X+5% over the next year!".

What you failed to realize was that earnings estimates will be decreased over the course of the year like clockwork, and eventually that yearly estimate that was too bullish will turn into a quarterly estimate that is too bearish.

The result of all this is the the graph above, with amateurs forecasting an extra 10% in quarterly revenue and extra 20% in EPS."

Do you mean "incentive to predict conservatively" rather than "incentive to under-predict"?

Sort of, "conservatively" would imply "keeping it safe" when predicting earnings, but I was really trying to raise the question of if the firms have anything to gain by deliberately making public predictions that are lower than they truly believe.

there are many studies that show that analysts are actually generally overly optimistic. there's no issue at play. the sample size here is one

This should be expected, bloggers are the 'market'. This is the expected result of the efficient market hypothesis. The market will beat estimates of professionals (or any one individual) most of the time. This is exactly why an index fund will outperform most mutual funds even before the fund manager takes his 2%.

The guys that can really guess the market usually start their own hedge fund. And they use ingenious methods, such as one hedge fund that orders satellite photos of Wal-Mart parking lots and estimates quarterly performance based in part on those photos.

I think you're confusing a lot of different things here. This is dealing with predicting quarterly results, there isn't any direct connection with the stock market.

Reply on news.ycombinator.com