On the risk of investing in stocks for the long run

Ask your investment, budget, and other money related questions here
jacob
Site Admin
Posts: 11152
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

On the risk of investing in stocks for the long run

Post by jacob » Wed Jul 17, 2019 3:25 pm

FYI: This paper reaches the opposite conclusion of what has become de facto investment dogma in the FIRE movement.

https://papers.ssrn.com/sol3/papers.cfm ... ct_id=5771
...the conventional wisdom may not apply to broad classes of individuals who face substantial human-capital risk early in their careers. For such individuals, the opposite policy may be optimal, i.e., to start out with a relatively low fraction of the investment portfolio in stocks
and increase it over time. Another critical determinant of the optimal investment in stocks is how close people are to some minimum “subsistence” level of consumption. People should be expected to insure against falling below such a level through their asset allocation policy.
The issue being that many in the FIRE movement rely on the results of long run statistics committing the "since the risk of failure is only 5%, it's not going to happen to me"-error of thinking. The case is made that this kind of thinking doesn't work on the individual practical level. What's interesting is that insofar one wants guarantees rather than chances, the strategy to follow as an extremely early retiree is the very opposite of what is generally believed.

daylen
Posts: 1048
Joined: Wed Dec 16, 2015 4:17 am

Re: On the risk of investing in stocks for the long run

Post by daylen » Wed Jul 17, 2019 4:35 pm

My thoughts on investing have never felt very sturdy. There seems to be an ongoing tension between static assumptions and the dynamic unknown. A frame that evaluates marginal cost/utility at a particular time given a particular allocation will always have a blindness/externality/randomness component. If this unknown has even a slight chance of ruining the entire portfolio, then all cost/utility optimization becomes meaningless. This appears to potentially be avoidable if life is assumed to be finite and preferences endure, because then the process could end in a non-ergodic state.

The internet and derivative construction is leading to a "standard" investing model that is growing along with increased systemic risk. It seems that strategy must evolve with the agent to avoid dissonance, but now the agents that build a frame with this information are being constrained into a standard of their own. So what can be done with this information? .. and with that information?

How can the required attention to play this game be justified if an agent does not enjoy the art of investing for its own sake? Perhaps all activities can be seen as a form of "investing", but then how do you account for your own limitations in converting between different forms of capital? At some point you hit a halting barrier and some form of faith is required to keep on calculating.

Just ranting. :)
Last edited by daylen on Wed Jul 17, 2019 4:41 pm, edited 1 time in total.

Seppia
Posts: 1008
Joined: Tue Aug 30, 2016 9:34 am
Location: Italy

Re: On the risk of investing in stocks for the long run

Post by Seppia » Wed Jul 17, 2019 4:36 pm

If someone on this board is willing to pay me 41.6% of their portfolio value in broad index funds against the possibility that it will underperform the risk free rate in the next 30 years, I’m a taker for all the money I can.

Seppia
Posts: 1008
Joined: Tue Aug 30, 2016 9:34 am
Location: Italy

Re: On the risk of investing in stocks for the long run

Post by Seppia » Wed Jul 17, 2019 4:47 pm

jacob wrote:
Wed Jul 17, 2019 3:25 pm
The issue being that many in the FIRE movement rely on the results of long run statistics committing the "since the risk of failure is only 5%, it's not going to happen to me"-error of thinking.
But this happens because people are usually very bad at math.
The same reason why if you explain the Martingale betting strategy to most people, they will think it’s bombproof.
jacob wrote:
Wed Jul 17, 2019 3:25 pm
insofar one wants guarantees rather than chances, the strategy to follow as an extremely early retiree is the very opposite of what is generally believed.
Is the point of the paper that, assuming $10k yearly expenses, one has a better probability of success accumulating $700.000 in TIPS vs $250.000 in stocks?

jacob
Site Admin
Posts: 11152
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: On the risk of investing in stocks for the long run

Post by jacob » Wed Jul 17, 2019 4:57 pm

@daylen - To borrow from Samuel Johnson, knowing that there's a risk of running out of money to pay food and rent concentrates the mind wonderfully.

@Seppia - And if the risk free rate (+inflation) does happen to outperform broad equity indexes in the next 30 years, will you still have the money to pay out? Therein lies the point of the paper. Basically, if you're offering to act as an insurance company that guarantees equity outperformance in the long run, what's your plan for covering the risk that it doesn't? I could also ask another way. Again, playing the insurance company, should you demand a higher premium for this guarantee (essentially a put option) for a 30 year horizon or a 60 year horizon?

Assuming $10k/yr in expenses... and a 70 year remaining lifespan, $700k in TIPS offers guaranteed success, whereas $250k in stocks offers a probability of a failure. Actual execution lies somewhere in between and depends on age. More comments in the paper.

jennypenny
Posts: 6212
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: On the risk of investing in stocks for the long run

Post by jennypenny » Wed Jul 17, 2019 6:21 pm

Worry-free Investing is worth a read if you haven't read it. I'm not sure if he's updated his approach or the book since I heard him speak a long time ago. He got burned early in his investing career which colored his viewpoint (too much at the time IIRC), but he was still a big influence on me and my contentment with 'having enough I can count on' (and not feeling badly when I miss out on some returns). I also remember him stressing that 'smart' people out-earned their needs and didn't need to gamble to cover them -- investing was to protect purchasing power, not expand it.

Anyway, it's been a long time so maybe I'm misremembering. It was a series of lectures ... I'm sure they're on youtube somewhere despite being over 20+ years ago.

User avatar
unemployable
Posts: 358
Joined: Mon Jan 08, 2018 11:36 am

Re: On the risk of investing in stocks for the long run

Post by unemployable » Wed Jul 17, 2019 6:27 pm

Only thing I see here is that as the time horizon increases, the cost of hedging against shortfall risk rises, approaching a limit of 1 at infinity.

As it should be. That's every risky asset.

But a single long-term hedge is cheaper than buying multiple shorter-term hedges over the same time period. A 10-year put, if you can find one, will cost less than five two-year puts or 10 one-year puts. In some cases the cumulative cost of the shorter-term puts exceeds 1. (An exception applies if vol decreases throughout the time horizon, but then that would violate the primary assumption of Black-Scholes, that vol is constant.)

Option prices structurally overvalue equity volatility for several reasons. That is, the volatility implicit in options prices tends to be higher than what is actually experienced, so that in the long run it is profitable to sell vol. The paper doesn't broach this topic, but does assume Black-Scholes is some sort of ideal, which it isn't. No one uses straight Black-Scholes anymore anyway, not since the crash of 87.
jacob wrote:
Wed Jul 17, 2019 4:57 pm
Assuming $10k/yr in expenses... and a 70 year remaining lifespan, $700k in TIPS offers guaranteed success
Not if the government defaults! And the US government has (FDR). And a default is always more likely over 70 years than over 30, so the same problem arises. In the case of TIPS they could also do something underhanded such as change the way CPI is calculated. Also a 70-year government bond doesn't exist, so you will face reinvestment risk at least twice.

Toska2
Posts: 265
Joined: Fri Nov 20, 2015 8:51 pm

Re: On the risk of investing in stocks for the long run

Post by Toska2 » Wed Jul 17, 2019 10:26 pm

Getting burned and not investing is worse than all bonds and getting a pittance.

So the article boils down to the lowest common denominator of investing personality.

Seppia
Posts: 1008
Joined: Tue Aug 30, 2016 9:34 am
Location: Italy

Re: On the risk of investing in stocks for the long run

Post by Seppia » Thu Jul 18, 2019 12:05 am

@jacob
My main criticism is better explained by unemployable, plus:
1- I actually do not plan on playing insurance company - I was being facetious. This said I would be more than willing to make a bet with anybody that (in spite of the terrible outlook of American equities) a world index fund will beat the risk free rate over 30 years.
We could do the bet with 1% of my net worth, assuming the world doesn’t collapse (if it does, we both lose).
Would you be willing to take this hypothetical bet?
IMHO, when the time horizon becomes much longer (es 100 years, but in some ways even 60) there are other risks in play, ie major paradigm shifts in how life on this planet works, but those would probably affect equally USA Tips and equities.

I do understand that $700.000 in tips guarantees success while $250.000 in stocks does not, but I feel we may be overlooking a small $450.000 detail :)

I kinda see the point of the paper, but I do not see what kind of really useful and actionable information I can get from it*.

It may be the first clickbaity titled paper I’ve ever seen though :)


*it can be useful if it instills a modicum is doubt into the 4% rule ayatollahs - but I’m far from being one of these.

classical_Liberal
Posts: 885
Joined: Sun Mar 20, 2016 6:05 am

Re: On the risk of investing in stocks for the long run

Post by classical_Liberal » Thu Jul 18, 2019 12:21 am

But as has been shown in the literature, the probability of a shortfall is a flawed
measure of risk because it completely ignores how large the potential shortfall might be.6
This is a hugely important piece of the puzzle people tend to miss. If someone runs historical simulations on a fixed dollar amount and are presented with two alternatives for investments. One of them provides a 5% chance of failure, but most of those failures are massive. The other, a 20% chance of failure, but all of those failures are only "near misses", ie portfolio failed to meet financial needs but only by small amounts. Which would you choose and why?

User avatar
unemployable
Posts: 358
Joined: Mon Jan 08, 2018 11:36 am

Re: On the risk of investing in stocks for the long run

Post by unemployable » Thu Jul 18, 2019 12:31 am

classical_Liberal wrote:
Thu Jul 18, 2019 12:21 am
This is a hugely important piece of the puzzle people tend to miss. If someone runs historical simulations on a fixed dollar amount and are presented with two alternatives for investments. One of them provides a 5% chance of failure, but most of those failures are massive. The other, a 20% chance of failure, but all of those failures are only "near misses", ie portfolio failed to meet financial needs but only by small amounts. Which would you choose and why?
"Failure" means the portfolio going to zero, right? And most ER models assume a withdrawal rate that is adjusted only nominally and only for inflation, not for portfolio performance. I mean you can model lowering your spending after a down year and some tools out there do just that, but if that's part of your policy statement I wouldn't call invoking it any sort of "failure".

I don't consider having less money than you started with at any point "failure", either. If you model for a 30-year horizon and start with 25 years of living expenses, then five years later have 21 years of expenses, you've gained a year. That's success!

What's the difference between a big failure and a little one? Zero is zero. You can't "owe" the stock market money.

classical_Liberal
Posts: 885
Joined: Sun Mar 20, 2016 6:05 am

Re: On the risk of investing in stocks for the long run

Post by classical_Liberal » Thu Jul 18, 2019 1:16 am

@unemployable
The difference is, IMO, huge. Having a 5% chance of hitting that zero point 20 years into a 50 year plan is fundamentally different than a 20% chance at year 45. Of course, each individual circumstance is different.

My point is that people take on extra risk simply to improve the chances of success, without looking at the failures. Why, and by how much, did it fail? That's the question more people should ask. That's how to move forward with confidence, identify and then mitigate the possibilities of failure.

Of course you can always just say, "save more money". I think that's become pretty common in the FIRE-sphere lately, the 4%-rule worship is dying down. But higher monetary goals often comes with it's own costs.

I still think absolute dollar amounts of spending is the most important factor in having the ability to mitigate poor financial performance with higher WR's. Simply because a few hundred bucks a month is so easy to come buy in Western countries. The question is whether or not that would have a material impact on an individuals draw-down.

Edit: Just to point out, the idea of retiring with less risky assets, then moving into more risk later is being discussed in the FIRE world. Although they are mostly discussing stock/bond percentages. Search bond tent or glidepath, it's also becoming the new buzz in mainstream financial planning.

slowtraveler
Posts: 767
Joined: Sun Jan 11, 2015 10:06 pm

Re: On the risk of investing in stocks for the long run

Post by slowtraveler » Thu Jul 18, 2019 6:20 am

You could live off dividends at 2% and have all the upside with less risk in terms of losing everything. The risk of government default is higher than all companies defaulting at once.

Inflation has been getting adjusted. They now account for substitution and deflation through technology progress. So if beef goes up in price, you substitute for chicken and the inflation weighs beef less.

There is no gaurantee of anything in life.

7Wannabe5
Posts: 4947
Joined: Fri Oct 18, 2013 9:03 am

Re: On the risk of investing in stocks for the long run

Post by 7Wannabe5 » Thu Jul 18, 2019 7:13 am

Insurance companies have to compete with other insurance companies for business, and there would almost certainly be another insurance company offering a lower rate based on Bayesian analysis applied towards creating larger risk pool rather than a hedge. IOW, based on methodology similar to that which allowed them to construct rates applicable to all chicken farmers to protect against poor production even though different breeds of chickens vary greatly in their laying rates.

Also, at some point in the analysis you have to touch base with the reality beyond the math likely attached to a situation such as "the overall valuation of the U.S. equity market declines by 20% each year for 5 years in a row." It is important to not be too reliant on what John Tukey referred to as "the fetish of objectivity."

jacob
Site Admin
Posts: 11152
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: On the risk of investing in stocks for the long run

Post by jacob » Thu Jul 18, 2019 7:31 am

I'm a dividend investor, but I will note that companies will cut their dividends insofar they do not have the cashflow or the ability to borrow more money to cover them. (See credit crisis where even blue chips cut their dividends.) This happens before they cut their preferred dividends which happen before they stop paying their [corporate] debt which happens before they default. Dividends are the most "junior" of responsibilities---right down there with job security and layoffs. One of the "bigger" issues that currently worry "serious people" is the increasing number of BBB/BBB- rated corporations in the US. Many have overextended themselves. Thanks, ZIRP.

Point being that diversification is no guarantee against a systemic failure.

This plays against the Fed which can essentially print their own money(*) or pursue other means of changing the public paper-balance, that is who holds cash and who holds bonds. During QE, it was changed so banks came to hold more cash and this cash was plowed into financial equity. If the cash had gone into capital (machines) or people ... we would have had inflation (and lots of it) instead.

Anyhow, countries with a sovereign currency are more likely to run the financial system into the ground with high inflation levels. Equity offers little protection against this. It is not an inflation hedge because at such a point the stockmarket is broken in the same way the "grocery"-market is broken: People don't know what anything costs anymore.

(*) Unlike say Greece, Puerto Rico, or Detroit.

The point that c_L makes is the important one. Within the traditional investing framework, presuming that we're looking at a 60 year horizon, there is no difference between a 90% drop after 10 years and spending oneself out of money and a 30% drop after 40 years that doesn't quite recover and running out of money in years 59. They both count as equal failures, but they are materially very different. The latter sucks but the former is catastrophic. Yet this is obscured when one statistically calculates the 5% failure rate ... Or even a 0% failure rate insofar the sampling space was inadequate.

Picture those Monte Carlo trajectories on firecalc. Now imagine a few lines just going south within a few years representing the mother of all financial crashes... they drop so fast that no amount of starting principal can recover them. Now historically (since 1870), this has not happened. But past returns are no guarantee of future returns.

In terms of "actionable" strategies, this suggests that the somewhat common or at least not uncommon strategy to have N months of cash at hand to ride out recessions is a smart idea. The 25% cash allocation of the permanent portfolio has a large cash position partially for this reason. It's inflation protected by gold (ha!) but also somewhat by the short end of the yield curve insofar interest rates can keep up. A portion of TIPS would serve the same role. What the paper literally suggest is to first CYA with TIPS, and then as you get richer (and closer to death) then add more risk with equity. This is the opposite of the usual idea (cf. the balanced portfolio approach) of assuming high risk when you're young and then taper it down as you get older. Taking on risk is fine insofar one can cover it with income from a job(*). Otherwise, the problem is that modern portfolio theory has deeply ingrained the idea that risk = reward ... and if you wait long enough, you shall surely get your reward => just take on as much risk as possible. Some people (in the FIRE movement---I have not seen anyone here) even go so far as to buy VTSAX on margin while waving their hands chanting "in the long run". Whereas in reality, higher risk = the chance of higher losses.

(*) But if you can't you have a problem. This is why I included that in the quote in the OP.

I'm bringing this up because the smart answer is somewhere in between the two extremes. It's hard to know what the extremes are, though, if one believes that one's own extreme is normal while being ignorant of the other extreme. Such is the current situation within FIRE.

User avatar
Dream of Freedom
Posts: 391
Joined: Wed Aug 29, 2012 5:58 pm
Location: Nebraska, US

Re: On the risk of investing in stocks for the long run

Post by Dream of Freedom » Thu Jul 18, 2019 7:34 am

Investing in bonds right now is a little like picking up pennies in front of a steam roller. If yields go up in the next decade or two your low yielding bonds would have a lower trading value. Tips may be a little insulated from that because presumably inflation would likely go with it. Still you could take a loss. The only reason I would want bonds right now is if I believed we will be in a deflationary environment soon.

Nomad
Posts: 170
Joined: Wed May 16, 2018 5:23 pm
Location: UK

Re: On the risk of investing in stocks for the long run

Post by Nomad » Thu Jul 18, 2019 7:45 am

I would say that personally, I do not have the right personality for investing in individual stocks.
If they go down unexpectedly, or up unexpectedly i panic and consider selling them...

7Wannabe5
Posts: 4947
Joined: Fri Oct 18, 2013 9:03 am

Re: On the risk of investing in stocks for the long run

Post by 7Wannabe5 » Thu Jul 18, 2019 8:12 am

So, the right answer is 80% timberland, meat rabbits, and guard dogs and 20% VTSAX on margin?

Seppia
Posts: 1008
Joined: Tue Aug 30, 2016 9:34 am
Location: Italy

Re: On the risk of investing in stocks for the long run

Post by Seppia » Thu Jul 18, 2019 8:28 am

jacob wrote:
Thu Jul 18, 2019 7:31 am
In terms of "actionable" strategies, this suggests that the somewhat common or at least not uncommon strategy to have N months of cash at hand to ride out recessions is a smart idea.
I agree 200%, and I'm slowly turning more defensive (have been for a little while now), via tilting the type of stocks I hold and increasing cash % in the AA.
I do believe we are heading for some rough waters.
My personal "unconventional" view is that bonds are even riskier than stocks right now, especially here in europe where bond yields are negative and stock valuations aren't as high as in the USA.

Ray Dalio makes a lot of interesting points here, and he goes even further by saying even cash will not be very safe.

https://www.linkedin.com/pulse/paradigm ... ublished=t

7Wannabe5
Posts: 4947
Joined: Fri Oct 18, 2013 9:03 am

Re: On the risk of investing in stocks for the long run

Post by 7Wannabe5 » Thu Jul 18, 2019 8:59 am

@Seppia:

Interesting article. What is your subjective judgment that Ray Dalio's subjective judgment is correct?


The other interesting thing to note about OP paper is that it also suggests that an older person such as me with short time frame until guaranteed subsistence income should take on more risk, which is what I already decided to do based on my recent reading.

Post Reply