Comments on: Regression to the Mean

By: dead hobo

dead hobo — Thu, 11 Dec 2008 17:22:30 +0000

Great chart, but it has one fatal flaw. It ignores fluctuations in the process mean. While it appears from this chart the market is a process that is in control and, overall, subject to random variation, the time period covered is too long for meaningful analysis.

I would submit that each crossing over or under the center line is in actuality a shift in the process mean and that each period should stand alone for analysis. This would aid the analysis substantially.

Now, the question of the day: Will the market cross the line and shift the mean to a lower point? Or will it jump upwards and continue above the line for a few more years? If it crosses the line and a recalculation of the line confirms the jump, kiss your sweet assets goodbye. If it remains above the line, then buy buy buy.

By: Mark E Hoffer

Mark E Hoffer — Thu, 11 Dec 2008 02:30:35 +0000

TrickStar Says:

December 10th, 2008 at 8:59 pm

TS,

Seriously, de nada, anytime.

By: Lance

Lance — Thu, 11 Dec 2008 02:01:09 +0000

I wrote about this at my place, but the most interesting thing about the second chart is that it does not imply that stocks are actually cheap. As I said there:

“if John Williams’ numbers are correct then stocks should be as low as they are because GDP and earnings have been horrible in inflation adjusted terms for a very long time. We have been in a recession for two decades similar to Japan. I don’t think that is true, but if so then stocks deserve every bit of the undervaluation they have experienced.”

My own opinion is that inflation is off, by by nowhere close to what Shadow Stats implies. Thus, the market is reasonably, but not spectacularly undervalued. That is pretty much what Jeremy Grantham says as well, and a minor adjustment to the first graph puts trend where he puts it as well.

By: TrickStar

TrickStar — Thu, 11 Dec 2008 01:59:40 +0000

@ Mark – That’s a fair point. I was excluding options from my example. Thanks for the link!

By: CC_in_Georgia

CC_in_Georgia — Thu, 11 Dec 2008 00:51:43 +0000

@StatArb:

Well said, sir. 100% correct.

The position of the regression line is solely dependent on the time series used. Therefore, the relationship (above or below trendline) between one particular data point and the regression line can vary wildly depending on what time period used.

By: JohnnyVee

JohnnyVee — Thu, 11 Dec 2008 00:22:23 +0000

The chart appears to show that in bad economic times that that bottom is about 50-80% under the mean. So we are no where near a bottom. If the mean would equal 850 today-by eye balling the chart- then only when the S&P is under 500, at the very least, should one start buying for the long term.

By: Mark E Hoffer

Mark E Hoffer — Thu, 11 Dec 2008 00:12:20 +0000

TS,

this: “That upward squiggly movement represent avg. returns of about 10% per year for a long, long time. Ain’t too many folks who have shown they can beat that consistently.”

is b/c ‘Cained Peep would rather be Wrong in groups, than Right alone..

any 7th-Grader could learn to do ‘Buy-Writes’ on:
http://finance.yahoo.com/q/bc?s=PM&t=1y
http://finance.yahoo.com/q/op?s=PM&m=2009-01

and walk home +10% per annum, with their eyes closed..

and, worse, that’s not even the best ex. (PM)

for some intro.. http://www.onn.tv/HomePage#close

By: TrickStar

TrickStar — Wed, 10 Dec 2008 23:53:50 +0000

1) Gee, those charts are great.
2) Buy and hold isn’t dead. If you look at the left side of either chart, you’ll see a little blue squiggly line start in the bottom left hand corner and make its squiggly way up to the upper right hand side of the chart. That upward squiggly movement represent avg. returns of about 10% per year for a long, long time. Ain’t too many folks who have shown they can beat that consistently.
3) Let’s be honest here. What really matters is which chart (of the two) the mutual fund and pension fund managers would use. Grantham likes the 10 year to determine fair values and mean reversions; I’m guessing he’d think the 135 years of data might distort things. If he’s using the 10-year, and you’re using the 135-year, you miss the train.

By: Steve Barry

Steve Barry — Wed, 10 Dec 2008 22:56:42 +0000

Intersting about the “bullish” alternative…what if we use the bullish alternative, but now assume rapid deflation going forward? The S&P would deflate as well, no?

By: Kent @ The Financial Philosopher

Kent @ The Financial Philosopher — Wed, 10 Dec 2008 22:53:59 +0000

Thanks for the various comments:

@DL: No, I did not move to 100% cash in October 2007. I had been slowly building cash since January 2007 (for myself and for my clients). At the time, most of my clients were unsure of my logic in the face of a healthy bull market but trusted the direction. Also, all new clients I have picked up since late 2007 and early 2008 were moved 100% to cash (depending on location and tax consequences) and set on a 12 month DCA schedule. That was not “timing the market” — it was simply logic, patience, simplicity and moderation.

@ Deltaverde: It sounds as if your “adviser” does not practice “risk management.” You can see my comment to DL for an example.

@ Bob in MA: “Fee-only” advisers are only paid by their clients and do not receive commissions; therefore, they are unbiased to products and the most client-centered adviser out there. If one was to seek an adviser, I recommend “fee-only.”

Most people can do as good or better investing on their own than with the vast majority of advisers. The problem is that our brains get in the way and the “average” investor often sells at low points and buys at high points.

With that said, however, commenters on this blog are leagues ahead of the “average” investor…