Do Models Always Fail When A Majority Applies Them?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

Do Models Always Fail When A Majority Applies Them?

Postby Scooter57 » Tue Mar 12, 2013 12:31 pm

I recently read a fascinating book, The Quants how a new breed of math whizzes conquered Wall Street and nearly destroyed it by Scott Patterson. It described the way that in the 1990s a bunch of math whizzes used modeling based on gambling theories to create very profitable hedge funds.

It then documented what happened in 2007 when financial events that the models predicted should be once in a lifetime events started happening every few days. The problem lay to a great extent in the fact that as soon as a critical number of these mathematically knowledgeable gamblers started using sophisticated tools that were based on the same models those models no longer would work.

I worry that this kind of effect may hit all of us who bought into the index funds theory in the 1980s and 90s when the concept was not popular as it is now. Now we have an environment with retail investors flooding into indexed, automated ETFs in numbers never before seen.

What impact will the presence in the market of all those indexed stock and bond ETFs and funds have on the theory that investing in indexed vehicles will beat choosing individual investments? But beyond that, what kind of unexpected devastation could the holding of so much equity and bonds in indexed investment vehicles wreak on the market in the next downturn? Is this what gets us to that 90% decrease in value we all assume is a once in a century kind of thing?

In 2008 when retail investors sold out of the market they sold mostly stocks and stock-picking funds. What happens when they sell the whole S&P 500?

Any thoughts about this you'd like to share?
Scooter57
 
Posts: 715
Joined: Thu Jan 24, 2013 9:20 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby larryswedroe » Tue Mar 12, 2013 3:29 pm

Suggest you read chapter 21 of my Wise Investing Made Simple which answers the question, what if everyone indexed.
Best wishes
Larry
larryswedroe
 
Posts: 12431
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Do Models Always Fail When A Majority Applies Them?

Postby Scooter57 » Tue Mar 12, 2013 3:31 pm

Thanks. I loved your bond book.
Scooter57
 
Posts: 715
Joined: Thu Jan 24, 2013 9:20 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby Rodc » Tue Mar 12, 2013 6:02 pm

Indexing does not use a model.

It is simply a way to get the average return of the market at low cost.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
 
Posts: 10519
Joined: Tue Jun 26, 2007 9:46 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby careytilden » Tue Mar 12, 2013 6:18 pm

Scooter57 wrote:What impact will the presence in the market of all those indexed stock and bond ETFs and funds have on the theory that investing in indexed vehicles will beat choosing individual investments?


I find this statement fascinating. Personally, I'm not using index based investments to beat any individual investments. I just want to grab the tiger by its tail and hope for a good return.
careytilden
 
Posts: 142
Joined: Tue Dec 01, 2009 2:43 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby avalpert » Tue Mar 12, 2013 7:41 pm

Your description of what the book says seems to conflate two distinct challenges faced by quant models. One is underestimating the likelihood and/or magnitude of fat tail events and the other is the feedback loops created by acting on the model itself.

Neither of these issues has the same impact on index-based investing as it does on quant models.
avalpert
 
Posts: 3412
Joined: Sat Mar 22, 2008 4:58 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby Scott S » Wed Mar 13, 2013 4:43 pm

The beauty of broadly-diversified Boglehead investing is that it doesn't matter what causes the market to go up or down -- you just rebalance when necessary, and things will take care of themselves if you save enough. :beer
My Plan: * Age-10 in bonds until I reach age 60, 50/50 thereafter. * Equity split: 50/50 US/Int'l, Bond split: 50/50 TBM/TIPS. * Everything over 2 months' expenses gets invested.
User avatar
Scott S
 
Posts: 1381
Joined: Mon Nov 24, 2008 3:28 am
Location: SW CR IA US NA PE

Re: Do Models Always Fail When A Majority Applies Them?

Postby Scooter57 » Fri Mar 22, 2013 9:16 pm

I went and read Chapter 21 of Larry's book, but did not find anything in it that answered the question I had posed here about the unintended consequences of a majority of investors adopting the indexing model. I also didn't find anything in that book I hadn't already read, far better expressed and with helpful illustrations, in William Bernstein's Four Pillars of Investing. I am very well aware of the arguments for indexing vs. market timing, which was all that Larry's chapter touched on. I'm investing using a simple index fund-based approach, a la Bernstein, which is why I spend time on this forum.

But the enthusiasm for indexing and the shift towards indexing by large institutional investors could very well lead to unexpected consequences for the market as a whole that might exacerbate the fluctuations of the market in a way that further damages the public's trust in the market, keeping investors out of the market, and stifling the ability of stock values to grow over the long term.

The reason people are enthusiastic about indexing is because they believe that by investing in an indexed total market approach they will end up with a decent gain over time. How many people here would stay the course if that meant going through 20 years of falling stock prices and dropping bond fund prices?

I'm not convinced this will happen, I only suggest that it could happen, because financial history is filled with examples of things that work very well for a while and then stop working when a critical mass of investors become aware of them.
Scooter57
 
Posts: 715
Joined: Thu Jan 24, 2013 9:20 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby leonard » Fri Mar 22, 2013 10:00 pm

No matter what the index does, by definition one will still get the index return, which is the point.

I am not sure what issue you are raising here.
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.
leonard
 
Posts: 4444
Joined: Wed Feb 21, 2007 11:56 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby Scooter57 » Sat Mar 23, 2013 7:03 pm

My point is that large amounts of indexing may destabilize the markets in ways that make "market return" unsatisfactory. That's because indexing links large blocks of assets so the market may move far more extremely than it did on the past.
Scooter57
 
Posts: 715
Joined: Thu Jan 24, 2013 9:20 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby gtwhitegold » Sat Mar 23, 2013 7:22 pm

The fact is that individual investors are a very small part of the market. So, if everyone followed the boglehead philosophy, they would continue to have minimal impact on the market. Buying index funds as money becomes available and selling them as needed to rebalance or to withdraw funds will not have a significant impact on the equities or fixed income markets. So, if a black swan event occurs it will be a very extreme example if it is caused by individual index investors. Hedge funds, large banks, and other entities controlling large amounts of fixed income and equities are most likely to effect the state of the market.

Allen
gtwhitegold
 
Posts: 57
Joined: Fri Sep 21, 2012 1:55 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby Kosmo » Sat Mar 23, 2013 7:40 pm

You use a model to predict what's going to happen to a system when it gets certain inputs. Under most circumstances those inputs are outside of the range of data which you used to create the model. And using the model, you're affecting the stock market and therefore changing the inputs, so there's feedback. If you use a model, you're basing your investment decisions (model output) on things like interest rates, currency valuations, stock prices, earnings reports, etc. (model inputs). As an index investor, I'm not using any model to make my decisions. I'm accepting that the stock market will do whatever it does, and I'll take an averaged return. The model may "fail" when the inputs become far out of range of the expected inputs. But using an index doesn't use a model, so it can't fail.
User avatar
Kosmo
 
Posts: 596
Joined: Wed Sep 05, 2012 11:54 am
Location: Philadelphia

Re: Do Models Always Fail When A Majority Applies Them?

Postby billjohnson » Sat Mar 23, 2013 8:08 pm

Scooter57 wrote:Do Models Always Fail When A Majority Applies Them?

Short Answer: Yes.
billjohnson
 
Posts: 585
Joined: Thu Nov 04, 2010 10:41 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby Noobvestor » Sat Mar 23, 2013 8:13 pm

Scooter57 wrote:The reason people are enthusiastic about indexing is because they believe that by investing in an indexed total market approach they [b]will end up with a decent gain over time[/b]. How many people here would stay the course if that meant going through 20 years of falling stock prices and dropping bond fund prices?


I think you added a false premise here (my bold above). I don't presume I will necessarily realize a substantial gain - I merely look at the alternatives (trying to actively trade, leaving my money in cash, gold or t-bills, etc... ) and conclude that indexing stocks/bonds is the most appealing option to me for expected risk-adjusted growth (or at least capital preservation). Could it lose you money over 20 years? Sure. Anything could, except, I suppose, long-term, positive-yield TIPS.
"In the absence of clarity, diversification is the only logical strategy" -= Larry Swedroe
User avatar
Noobvestor
 
Posts: 3987
Joined: Mon Aug 23, 2010 1:09 am

Re: Do Models Always Fail When A Majority Applies Them?

Postby nedsaid » Sat Mar 23, 2013 9:14 pm

I don't think you have to worry about the index strategy failing. The reason is that indexing is a boring strategy. The people that crave exitement will not stick to index investing. They will move on to something more exciting. The markets will never be 100 percent indexed.
A fool and his money are good for business.
User avatar
nedsaid
 
Posts: 2763
Joined: Fri Nov 23, 2012 12:33 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby Phineas J. Whoopee » Sun Mar 24, 2013 1:37 pm

Scooter57 wrote:...
I worry that this kind of [bad] effect may hit all of us who bought into the index funds theory in the 1980s and 90s when the concept was not popular as it is now. Now we have an environment with retail investors flooding into indexed, automated ETFs in numbers never before seen.
...
In 2008 when retail investors sold out of the market they sold mostly stocks and stock-picking funds. What happens when they sell the whole S&P 500?

Any thoughts about this you'd like to share?

Hi Scooter,

The question about whether indexing sows its own doom comes up regularly here. In addition to points made by previous posters, if a great majority of the value of equity markets was held by cap-weight indexers, the pricing mechanism would begin to lose efficiency. That would open opportunities for arbetrageurs. People being as they are, some would take advantage of the chance, which would tend to move pricing back toward greater efficiency.

In a nutshell, the point beyond which investors indexing would, in and of itself, distort markets would not be reached. Instead there would become an equilibrium between indexers and those extracting arbitrage profits.

With respect to the effect of indexed market participants leaving all at once: to the extent they did so when not indexed, some held one fund, some another, and the net total is the total market. That skittish investors acting against their own interests might now use more- rather than less-efficient vehicles would not change the impact.

PJW
User avatar
Phineas J. Whoopee
 
Posts: 3222
Joined: Sun Dec 18, 2011 6:18 pm

Re: Do Models Always Fail When A Majority Applies Them?

Postby leonard » Sun Mar 24, 2013 1:56 pm

Scooter57 wrote:My point is that large amounts of indexing may destabilize the markets in ways that make "market return" unsatisfactory. That's because indexing links large blocks of assets so the market may move far more extremely than it did on the past.


That would have been a good topic sentence.

The reality is that you don't need many companies executing arbitrage to keep all the returns in line across financial instruments. I forget the exact percentage - but you only require a small percentage of the transaction to be arbitrage transactions - in order to keep the risk-return of indexes and all the various financial instruments tied to them in order.
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.
leonard
 
Posts: 4444
Joined: Wed Feb 21, 2007 11:56 am


Return to Investing - Theory, News & General

Who is online

Users browsing this forum: Yahoo [Bot], Zedon and 28 guests