Fat Tail Risk vs Asset Allocation

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

Fat Tail Risk vs Asset Allocation

Postby 1210sda » Tue Mar 15, 2011 9:47 pm

Does fat tail risk decrease as you increase your allocation to bonds (intermediate term)?? Alternatively, does risk increase as you increase your stock allocation.?

Ceteris Paribus.......

1210
User avatar
1210sda
 
Posts: 787
Joined: Wed Feb 28, 2007 9:31 am

Postby craigr » Tue Mar 15, 2011 10:02 pm

No. Bonds have their own risks with inflation. A black swan inflation-type event with the US dollar would be very bad for a heavy bond allocation.

All investments have risks. It's just a matter of understanding what the risks are and diversifying against them as best you can in case they manifest.
User avatar
craigr
 
Posts: 2674
Joined: Tue Mar 13, 2007 7:54 pm

Re: Fat Tail Risk vs Asset Allocation

Postby pkcrafter » Wed Mar 16, 2011 12:15 pm

1210sda wrote:Does fat tail risk decrease as you increase your allocation to bonds (intermediate term)?? Alternatively, does risk increase as you increase your stock allocation.?

Ceteris Paribus.......

1210


Fat tails means there are more dramatic events than a normal distribution will predict. These events occur more often than predicted and can be much more severe than anything in recent history. For investors, the great fear is not the frequency, but the magnitude of a drawdown. We frequently use a rule of thumb that says you might lose 50% of what your portfolio allocation to equity is, but a black swan event can easily exceed this rule of thumb. As to your question, yes, lower allocations to stock will reduce potential drawdown. But, as craigr suggests, there are potential black swans in all phases of investing.

Some investors, well aware of the nature of risk, can tolerate a substantial loss, but it's clear that many still bail when the market is down 50%. Why? Because once they are down 50%, they fear a deeper loss, which they cannot tolerate. That is the real threat. So, what happens to the investor who can grit his teeth and hold and the market does drop below a 50% loss. AT 60%, gains needed to recover begin to increase exponentially. 150% gain is needed to recover a 60% loss. An 80% loss--a NASDAQ black swan event in 2000-2002--requires a 400% gain. So, it may not really be a question of whether the market can break you or not; the question may be how much is it going to take for the market to break you.

As to your question, lower allocations to stock will reduce potential magnitude of a drawdown. But, as craigr suggests, there are potential black swans in all phases of investing.

Here's an article on dealing with black swans

http://www.forbes.com/2009/10/14/gerstein-fisher-nyu-personal-finance-fat-tails.html

A link to online book chapter on asset allocation and risk

http://investingroadmap.wordpress.com/2011/02/25/205/


Paul
Last edited by pkcrafter on Wed Mar 16, 2011 4:30 pm, edited 2 times in total.
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.
pkcrafter
 
Posts: 8162
Joined: Sun Mar 04, 2007 1:19 pm
Location: CA

Postby 1210sda » Wed Mar 16, 2011 12:26 pm

Paul,

I understand there are risks in all aspects of investing.

I was just wondering if there was a greater risk of a fat tail with a larger allocation to stocks.

For example, If you have a 100% equity portfolio with a standard deviation of 20 versus a 50/50 portfolio with a standard deviation of 8, wouldn't it seem that the fat tail impact would be less (perhaps much less) in the 50/50 portfolio ?

1210
User avatar
1210sda
 
Posts: 787
Joined: Wed Feb 28, 2007 9:31 am

Postby pauliec84 » Wed Mar 16, 2011 12:34 pm

Stocks are going to give you a greater exposure to FAT tails. This is the idea of Larry Swedroe portfolio which invests in riskier stocks, but puts a bigger allocation in TIPs.

Plus look at at this way. With 2 Types of Assets (50/50 portfolio), there are some black swans events that would only affect one of the asset types. I you diversify internationally (I see many black swans as political risk events), you could further reduce your black swan exposure.

With that said there are some "black swan" events, that would wipe out everything. Think of the US defaulting, Treasury's would crash, and I would imagine the equity market would too.
pauliec84
 
Posts: 342
Joined: Wed Feb 09, 2011 12:29 pm

Postby TheEternalVortex » Wed Mar 16, 2011 12:36 pm

TIPS are about as safe as you can get, so if you increase your allocation to TIPS you are decreasing the fat tail risk (of course this also includes the "risk" of high positive returns).
TheEternalVortex
 
Posts: 2420
Joined: Tue Feb 27, 2007 10:17 pm
Location: San Jose, CA

Postby dbr » Wed Mar 16, 2011 12:41 pm

1210sda wrote:Paul,

I understand there are risks in all aspects of investing.

I was just wondering if there was a greater risk of a fat tail with a larger allocation to stocks.

For example, If you have a 100% equity portfolio with a standard deviation of 20 versus a 50/50 portfolio with a standard deviation of 8, wouldn't it seem that the fat tail impact would be less (perhaps much less) in the 50/50 portfolio ?

1210


If by fat tail you mean that the incidence of stock returns away from the average being greater than described by a normal distribution, then obviously by simple arithmetic the impact of those incidences is diluted by holding (more) bonds.

A point that is being made is that bonds have their own risks, risks that could include rare but extreme cases. It does take an extreme case for fat tails in bond returns to have a bigger impact than the analogous fat tails in stock returns.

I don't consider in my mind that fat tails and black swans are talking about the same thing at all. The departures from the statistics of normal distributions referred to as fat tails are common and well known. Black swans are supposed to be events that are beyond reasonable attempts at statistical prediction. Perhaps some people imagine a ground in between, at the far extremes of non-normal distributions but still comprehensible in principle as probabilities that can estimated.
dbr
 
Posts: 14226
Joined: Sun Mar 04, 2007 10:50 am

Postby Rodc » Wed Mar 16, 2011 12:44 pm

1210sda wrote:Paul,

I understand there are risks in all aspects of investing.

I was just wondering if there was a greater risk of a fat tail with a larger allocation to stocks.

For example, If you have a 100% equity portfolio with a standard deviation of 20 versus a 50/50 portfolio with a standard deviation of 8, wouldn't it seem that the fat tail impact would be less (perhaps much less) in the 50/50 portfolio ?

1210


To the extent that Bonds are safer than Stocks the answer seems to me to be yes.

Sure, no allocation is without risk. It does not follow that all portfolios have the same degree of risk.

"Fat tail" often has the direct meaning that the underlying distribution is not as nice as a Gaussian (Normal) distribution, but I don't think that is quite your question, so I would drop that word.

But even sticking with Gaussian plus black swans, if bonds are less subject to black swans than stocks, a bond heavy portfolio is less subject to black swans than a stock heavy one.

Unless I misunderstand the question, this is totally accepted wisdom and underlies such rules of thumb as "age in bonds" (among others). It also means that simple Gaussian (mean and standard deviation) examples like the one you gave miss the important issue here).

Mean and standard deviation examples offer only very simple and incomplete understanding, which is what I think some of the first replies are getting at. (though if I am wrong the posters will let me know. :) )

Added, while typing my post several more appeared above, and my reference to first replies does not include them.
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: 10201
Joined: Tue Jun 26, 2007 10:46 am

Re: Fat Tail Risk vs Asset Allocation

Postby YDNAL » Wed Mar 16, 2011 1:01 pm

1210sda wrote:Does fat tail risk decrease as you increase your allocation to bonds (intermediate term)?? Alternatively, does risk increase as you increase your stock allocation.?
For example, If you have a 100% equity portfolio with a standard deviation of 20 versus a 50/50 portfolio with a standard deviation of 8, wouldn't it seem that the fat tail impact would be less (perhaps much less) in the 50/50 portfolio ?
1210,

Are you just throwing buzz-words around? :)

Stocks have risks. Bonds have risks.

Certain events - like a global financial meltdown (08/09) - can severely impact either... meaning, the normal distribution of risk is abnormal (fat tail). Of course, the lower the allocation to riskier investments (Stocks), the lower the impact those certain events can cause.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
YDNAL
 
Posts: 13567
Joined: Tue Apr 10, 2007 5:04 pm
Location: Biscayne Bay

Postby 1210sda » Wed Mar 16, 2011 1:46 pm

Landy,

"Are you just throwing buzz-words around? "

I don't understand your question. Am I not allowed to ask a question? Does it have to be worded in such a way as to meet your approval ?

1210
User avatar
1210sda
 
Posts: 787
Joined: Wed Feb 28, 2007 9:31 am

Re: Fat Tail Risk vs Asset Allocation

Postby YDNAL » Wed Mar 16, 2011 2:29 pm

1210sda wrote:I don't understand your question. Am I not allowed to ask a question? Does it have to be worded in such a way as to meet your approval ?

1210

No comment... dead horse.
Image
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
YDNAL
 
Posts: 13567
Joined: Tue Apr 10, 2007 5:04 pm
Location: Biscayne Bay

Postby pkcrafter » Wed Mar 16, 2011 4:40 pm

sda wrote:
For example, If you have a 100% equity portfolio with a standard deviation of 20 versus a 50/50 portfolio with a standard deviation of 8, wouldn't it seem that the fat tail impact would be less (perhaps much less) in the 50/50 portfolio ?

The answer is yes, a 50/50 portfolio will not have the drawdown of a 100% stock portfolio, but the reason is not the standard deviation--it's the equity exposure itself. In a panic drawdown the SD of all equity asset classes tend to approach 1, so your primary protection is how much exposure your portfolio has to equities.


Paul
When times are good, investors tend to forget about risk and focus on opportunity. When times are bad, investors tend to forget about opportunity and focus on risk.
pkcrafter
 
Posts: 8162
Joined: Sun Mar 04, 2007 1:19 pm
Location: CA


Return to Investing - Theory, News & General

Who is online

Users browsing this forum: CP4641, Google [Bot], happyisland, jj2323us, jjbiv, linenfort, The Wizard, vectorizer and 62 guests