Author Topic: trinity study in another dimension  (Read 2451 times)

solon

  • Handlebar Stache
  • *****
  • Posts: 1335
  • Age: 1817
  • Location: CO
trinity study in another dimension
« on: May 04, 2018, 08:18:00 AM »
This guy thought the classic Trinity study was based on too few observations - there are not enough 30-year periods to give meaningful results. So he simulated 1000 30-year periods. The results confirm the original Trinity study, but back it up with a lot more data.

https://rockstarfinance.com/trinity-study-another-dimension/

sol

  • Walrus Stache
  • *******
  • Posts: 6819
  • Age: 41
  • Location: Pacific Northwest
Re: trinity study in another dimension
« Reply #1 on: May 04, 2018, 08:33:11 AM »
This guy thought the classic Trinity study was based on too few observations - there are not enough 30-year periods to give meaningful results. So he simulated 1000 30-year periods. The results confirm the original Trinity study, but back it up with a lot more data.

https://rockstarfinance.com/trinity-study-another-dimension/

Without even clicking the link, let me guess the flavor of chicanery.

Did he reshuffle every year's historical stock market performance into a quasi-random string of returns that is 3,000 years long?  Did he call that a Monte Carlo simulation?

That's the usual (deeply flawed) methodology for these sorts of people.  Even Vanguard has done it.  It's wrong.  So so very very wrong.

Why is it wrong?  Because stock market performance is not an independent variable.  This year's return depends strongly on returns over the previous years, and the distribution of possible outcomes is not nearly so tidy as we would like to believe.  Sure, the US market has historically returned about 10% per year, there are instances of stock markets going to zero (for example when nations collapse and disappear) that do not show up in this sort of Monte Carlo analysis.

The reality of this situation is that the tea leaves are read in broad economic trends, not random variables.  Is your nation being invaded?  Is there a global pandemic, or an immigration crisis (by which I mean a lack of it)?  Is your economy dominated by a single firm (ala Samsung) that could go bankrupt and take your entire national economy with it?  Do you have kindergartners in charge of your monetary policy?  Your trade policy?  Is your monarch in failing health and being usurped by an upstart with unproven new ideas?  What is the likelihood of one or more of these types of stressors showing up in the future of your stock market?  Because these are the sorts of trends that can be used to predict future returns and they require knowledge of history and politics, not statistics.  If you just repeat the historical record of market performance over and over again you're not actually bringing any new predictive powers to bear on the problem, you're just filling in the gaps between points on the existing graph.

OurTown

  • Pencil Stache
  • ****
  • Posts: 862
  • Age: 48
  • Location: Tennessee
Re: trinity study in another dimension
« Reply #2 on: May 04, 2018, 08:57:33 AM »
No records on the paleolithic bead trade, sorry.

Xlar

  • Stubble
  • **
  • Posts: 124
Re: trinity study in another dimension
« Reply #3 on: May 04, 2018, 10:28:28 AM »
This guy thought the classic Trinity study was based on too few observations - there are not enough 30-year periods to give meaningful results. So he simulated 1000 30-year periods. The results confirm the original Trinity study, but back it up with a lot more data.

https://rockstarfinance.com/trinity-study-another-dimension/

Without even clicking the link, let me guess the flavor of chicanery.

Did he reshuffle every year's historical stock market performance into a quasi-random string of returns that is 3,000 years long?  Did he call that a Monte Carlo simulation?

That's the usual (deeply flawed) methodology for these sorts of people.  Even Vanguard has done it.  It's wrong.  So so very very wrong.

Why is it wrong?  Because stock market performance is not an independent variable.  This year's return depends strongly on returns over the previous years, and the distribution of possible outcomes is not nearly so tidy as we would like to believe.  Sure, the US market has historically returned about 10% per year, there are instances of stock markets going to zero (for example when nations collapse and disappear) that do not show up in this sort of Monte Carlo analysis.

The reality of this situation is that the tea leaves are read in broad economic trends, not random variables.  Is your nation being invaded?  Is there a global pandemic, or an immigration crisis (by which I mean a lack of it)?  Is your economy dominated by a single firm (ala Samsung) that could go bankrupt and take your entire national economy with it?  Do you have kindergartners in charge of your monetary policy?  Your trade policy?  Is your monarch in failing health and being usurped by an upstart with unproven new ideas?  What is the likelihood of one or more of these types of stressors showing up in the future of your stock market?  Because these are the sorts of trends that can be used to predict future returns and they require knowledge of history and politics, not statistics.  If you just repeat the historical record of market performance over and over again you're not actually bringing any new predictive powers to bear on the problem, you're just filling in the gaps between points on the existing graph.

This is very insightful, thank you!

NoStacheOhio

  • Handlebar Stache
  • *****
  • Posts: 2152
  • Location: Cleveland
Re: trinity study in another dimension
« Reply #4 on: May 04, 2018, 01:32:22 PM »
Also, a "year" is (basically) arbitrary, and generally shorter than the time it takes for a given economic trend to play out.

sol

  • Walrus Stache
  • *******
  • Posts: 6819
  • Age: 41
  • Location: Pacific Northwest
Re: trinity study in another dimension
« Reply #5 on: May 04, 2018, 01:49:32 PM »
Also, a "year" is (basically) arbitrary, and generally shorter than the time it takes for a given economic trend to play out.

The time period is basically arbitrary.  We've seen trinity results reproduced with monthly performance results, and now I'm waiting for someone to publish it with the minute by minute market index history all scrambled up and randomized for another thousands years.

The underlying motivation of these efforts is to calculate how long you need to randomize the historical data into the future until you get all of the worst years (months, minutes) side by side, and then say "See!  I've discovered a catastrophic future where everything collapses!"  Even vanguard did this, publishing a simulation with an arbitrary sequence like1931 happens right after 2008 right after 1907 right after 1974, and then claiming that the market could realistically be expected to decline 96% based in the historical record.  They completely ignore the economic fundamentals that underpinned these individual bad years, and which explain why they didn't (and arguably can't) actually happen back to back.

Monte Carlo simulations are a powerful tool for understanding nonlinear responses of data parameterization, but I don't think they have any place in historical stock market analyses.

aspiringnomad

  • Pencil Stache
  • ****
  • Posts: 713
Re: trinity study in another dimension
« Reply #6 on: May 04, 2018, 02:03:38 PM »
Do you have kindergartners in charge of your ... trade policy?

Uh oh...

Eric

  • Magnum Stache
  • ******
  • Posts: 3903
  • Location: On my bike
Re: trinity study in another dimension
« Reply #7 on: May 04, 2018, 02:07:29 PM »
I'm not a fan of that guy's writing style.  That's a lot of words to simply say "I ran a Monte Carlo sim instead".

Would we all like more data points?  Sure, of course.  But they simply don't exist.  This fake creation of extra "data" is way worse than using the smaller sample size.  As Sol rightly points out, stock market performance does not exist in a vacuum and returns are not independent events.

I also don't really understand the fascination with finding non-overlapping periods.  The fact that a retirement period from 1946-1975 and 1972-2001 both contain the year 1973 hardly makes them similar.

So then after all of that, despite the fact that his Monte Carlo sim showed a 1.8% WR at 100% or a 3.8% WR at 95% success rates, he decides that he's comfortable with a WR greater than 4%.  Great, so what was the point of that?  He could've just used the Trinity Study data.

solon

  • Handlebar Stache
  • *****
  • Posts: 1335
  • Age: 1817
  • Location: CO
Re: trinity study in another dimension
« Reply #8 on: May 04, 2018, 03:19:27 PM »
The big thing I took away from this is that the Trinity results turn out pretty much on target. In fact, Trinity might even be a little pessimistic. In the linked study, it looks like you'd be safe with a 5% withdrawal rate.

And I see you guys have thought about this a lot more than I have. I never heard of a Monte Carlo simulation before this, but I guess it's a pretty common exercise?

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1037
Re: trinity study in another dimension
« Reply #9 on: May 04, 2018, 04:06:02 PM »
I don't think there's anything wrong with Monte Carlo simulations per se. It's a valid statistical technique.

I wish the author had more clearly explained why they made the choice to put each number back in the hat for the next draw. Mean reversion was repeated a few times, so my guess is this strategy is meant to prevent a mean reversion effect and keep the sample random. Had they only allowed 1929 or 2008 to be selected once, the remaining numbers in the hat would have a higher average and tend to bring up the ROI for subsequent years. As they did it, they can call it random FWIW.

The real problem IMO is only using U.S. historical data. The rise of stock markets coincided with the rise of the U.S. as the world's biggest economic superpower. Results varied in the stock markets of other countries. Mexico, for instance, started the 20th century in a similar economic position as the U.S. but cultural and governance factors led to a slight gap in growth that became an enormous economic difference over time. In 1900, you would not have known which market to invest in. Yet, Mexico's stock market is not included in the analysis. Nor is Japan's, Argentina's, Turkey's, Germany's, France's, Brazil's, or the UK's. The SWR in those countries is anyone's guess.

This matters because the U.S.' economic acheivements in the 20th century either may or may not be repeated. Lots has changed since those results were made: Glass-Stegal was repealed, the debt has skyrocketed, the population has aged, government and the finance industry are bigger parts of the economy, and certain easy-to-obtain resources have been depleted. The quality of our governance and levels of corruption are probably different too. Thus, the future U.S. is best extrapolated from worldwide results. Then do the monte carlo analysis.

Eric

  • Magnum Stache
  • ******
  • Posts: 3903
  • Location: On my bike
Re: trinity study in another dimension
« Reply #10 on: May 04, 2018, 04:44:51 PM »
The real problem IMO is only using U.S. historical data. The rise of stock markets coincided with the rise of the U.S. as the world's biggest economic superpower. Results varied in the stock markets of other countries. Mexico, for instance, started the 20th century in a similar economic position as the U.S. but cultural and governance factors led to a slight gap in growth that became an enormous economic difference over time. In 1900, you would not have known which market to invest in. Yet, Mexico's stock market is not included in the analysis. Nor is Japan's, Argentina's, Turkey's, Germany's, France's, Brazil's, or the UK's. The SWR in those countries is anyone's guess.

There is research on it.  Here's a paper from Pfau that talks about it.

https://www.advisorperspectives.com/articles/2014/03/04/does-international-diversification-improve-safe-withdrawal-rates

The general trend is that if you ignore scenarios where you lost (or were decimated by) a major war, then it's somewhere around 3% in most countries.  Which is lower than the US-centric 4%, but of course, it's also not a great premise to run a simulation where someone invests their entire portfolio in only companies from The Netherlands, ignoring all other markets.  So it's debatable how useful country specific data is, or what the specific SWR for France is, because that's not good investor behavior to begin with.

Now he does attempt to apply global diversification to these SWRs using a 50/50 AA.  Conclusion:
Quote from: Pfau
When aggregated across countries, the success rates for a 4% withdrawal rate with 50/50 portfolios over 30 years rose from 65.7% with only domestic assets to 78.3% with globally diversified assets.

So basically 4% is knocked down from a US 95% success rate to a Global 78% success rate.  And then you can likely apply an upwards adjustment for holding more stocks instead of 50/50, so that's still pretty good odds in my mind. 


maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #11 on: May 04, 2018, 05:17:19 PM »
I don't think there's anything wrong with Monte Carlo simulations per se. It's a valid statistical technique.

It's a valid statistical technique for certain problems. Like any statistical technique if you apply it to problems that violate its assumptions you get misleading output. One of the key assumptions of monte carlo analysis is the absence of correlation between sampling units.

In this case the true historical record of the stock market exhibits substantial autocorrelation from year to year. Bad years are more likely to follow other bad years than good years and vice versa.

What the means is that you see a greater variation in total return from real 30 year periods than from synthetic 30 year periods constructed from a random ordering of random years selected from the exact same dataset. Events like the great depression (-13% inflation adjusted returns returns, followed by -21% returns the next year, -42% returns the year after that and -5% the year after that) or the 1970s (-17% inflation adjusted, then -5%, then 7%, then 12%, then -24% then -37%) are far rarer in monte carlo generated 30 year intervals data than they are in 30 year intervals drawn from the historical stock market record.

Since failures are a result of the rare periods where total 30 year returns are substantially lower than average, the reduction in variance in the synthetic data means that it produces falsely optimistic forecasts for both failure rates and the maximum sustainable safe withdrawal rates.

This, in turn produces reactions from people who haven't dug into monte carlo simulations and the situations where they are and aren't appropriate like this one (not to pick on you solon):

The big thing I took away from this is that the Trinity results turn out pretty much on target. In fact, Trinity might even be a little pessimistic. In the linked study, it looks like you'd be safe with a 5% withdrawal rate.

TL;DR version: What @sol said.

wbarnett

  • 5 O'Clock Shadow
  • *
  • Posts: 2
  • Location: Denver, CO, USA
Re: trinity study in another dimension
« Reply #12 on: May 05, 2018, 10:02:14 AM »
The replies above are spot-on. To the extent that the autocorrelation is not included in the simulation, the output from the simulation is pretty meaningless.

I'm a statistician; I'm not sure why these types of articles don't build in some correlation. It's fairly simple to fit an AR process to periods of returns and simulate from that. Or you could make it a Bayesian hierarchical model and put a reasonable distribution on the correlation. The fact that the analysis came from an actuary is a little depressing. Surely he/she is capable of doing something more sophisticated than randomly sampling returns with replacement...

sol

  • Walrus Stache
  • *******
  • Posts: 6819
  • Age: 41
  • Location: Pacific Northwest
Re: trinity study in another dimension
« Reply #13 on: May 05, 2018, 10:16:05 AM »
The fact that the analysis came from an actuary is a little depressing. Surely he/she is capable of doing something more sophisticated than randomly sampling returns with replacement...

It's only depressing if you think the point of the exercise is to generate "valid" results rather than "newsworthy" results.  When viewed through the lens of the practitioners motivations, it makes more sense that he might choose deliberately bad methods.

In this case I think he blew it, though.  Why choose an invalid model and then use it to declare the existing model sound?  If you're going to torture your data, at least torture it into confessing something interesting.

sol

  • Walrus Stache
  • *******
  • Posts: 6819
  • Age: 41
  • Location: Pacific Northwest
Re: trinity study in another dimension
« Reply #14 on: May 05, 2018, 11:06:06 AM »
A study of corporate balance sheets would reveal that the mix of assets that make up the value creation engines of the real economy, balance sheet items, like cash, property, patents, production facilities, commodity reserves, etc. are shifting gradually from being mostly dominated by hard assets like land and commodities (e.g. like oil), to an increasing share of asset value being driven by information, process knowledge, and unique people skills. 
... 
Information based companies and political entities active in the space (e.g. russian intelligence) are becoming huge political power brokers

I agree that the economy is changing dramatically from the one that generated our stock history charts, and that future returns could be very different (better or worse) as a result.

This is exactly the kind of analysis that I think is more useful than a Monte Carlo of return histories.  You need to try to understand the arc of history to predict the future, not just reshuffle the past.

With that said, history hasn't exactly been stagnant either, and yet market returns have been consistently positive for a century, with little wiggles in between, as human society finally pulls itself out of the bronze age.  My primary concern with this approach is that we're nearsighted when we only look at the history of stock returns during the rise of capitalism, while ignoring the larger economic shifts that have characterized the centuries.  Feudal Lords prospered for a century too, but ultimately their entire economic model was made obsolete.  Ditto for industrialists, and slavers, and mercantilists, and seafaring traders, and the French aristocracy, and the Pharaohs of Egypt.  I suspect that a hundred years of stock market profits are fine for estimating 30 years into the future, but eventually this too shall pass. 
« Last Edit: May 05, 2018, 08:49:20 PM by sol »

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #15 on: May 05, 2018, 01:15:14 PM »
Feudal Lords prospered for a century too, but ultimately their entire economic model was made obsolete.  Ditto for industrialists, and slavers, and mercantilists, and seafaring traders, and the French aristocracy, and the Pharaohs of Egypt.  I suspect that a hundred years of stock market profits are fine for estimating 30 years into the future, but eventually this to shall pass.

If you look at former upper (or ruling) classes or castes like the ones you list, their ends came in one of two ways:

1) Society changes and their role went away, but the individual people involved tended to still see their standards of living increase, it was just the the rest of society caught up to some extent, combined with whatever made the group distinctive fade away. For example the descendants of english nobility, or of 18th century industrialists like Rockefeller and Carnegie are still doing reasonably well for themselves today. And I'm guessing few of them would give up modern conveniences to go back to a world where the gap between their lifestyles and the median households was much larger.

2) Violent revolution or invasion. The descendants of french nobility or slaveholders in the american southeast very rapidly lost their economic and social positions (and sometimes their lives), and if they had surviving descendants they lived through some very undesirable experiences. The key here would seem to be to avoid being visible or socially identified with the ruling class, and also be aware and willing to leave for another part of the globe if things get too bad. (But how does one definite "too bad"? That's the million dollar question).

...I suppose technically we could add a third case, which would be the economic systems of civilizations that ended up in that Jared Diamond Book "Collapse." But if we're living through the equivalent of the last years of the Viking settlements on Greenland, or of Easter Island, it really doesn't matter what we do as individuals, we're in for a bad end, so I mostly focus on the first two cases.

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1037
Re: trinity study in another dimension
« Reply #16 on: May 06, 2018, 08:53:14 PM »
I agree that the economy is changing dramatically from the one that generated our stock history charts, and that future returns could be very different (better or worse) as a result.

This is exactly the kind of analysis that I think is more useful than a Monte Carlo of return histories.  You need to try to understand the arc of history to predict the future, not just reshuffle the past.

Be careful, Sol. Narrative explanations are easy for anyone to produce. They're basically a Rorchach test, and investors who place bets based on a "thesis" are often burned by their own reasoning. Remember the meme often posted here of the stock market zooming up as all sorts of risks loom or are realized?

The stories we write for ourselves are motivated cognition, done under the influence of limited or curated information about the past. E.g. You might have learned in school why the Great Depression occurred, but was your lesson accurate? All of us have an understanding of the Great Financial Crisis that we obtained from media reports. Are they accurate? Are complexities ever simplified for public consumption to the extent the meaning of the information is changed?

To this day, there are goldbugs waiting with their collector coins for the great hyperinflation that Fox News assured them would come within months... back in 2009. They'd be millionaires had they not allowed Glenn Beck to influence their worldview, and invested in equities instead.

The old "buy-and-hold a highly diversified equity portfolio" advice has worked fairly well for many decades. That is not a reason to be comfortable with it, because the future/present doesn't always resemble the past. However it does suggest that a method-outcome link is supported by enough durable factors to overwhelm the influence of significant noise. The monte carlo analysis, methodological concerns aside, suggests that the 4% rule might be another highly durable method-outcome link. That link would break with the impact of a mile-wide asteroid, but at some point the data has to satisfy us with our decisions, even if it cannot rule out a change of the rules.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #17 on: May 06, 2018, 09:08:27 PM »
The monte carlo analysis, methodological concerns aside, suggests that the 4% rule might be another highly durable method-outcome link. That link would break with the impact of a mile-wide asteroid, but at some point the data has to satisfy us with our decisions, even if it cannot rule out a change of the rules.

And assassination aside, how did you like the play Mrs. Lincoln?

The methodological concerns are a fatal flaw. This particular analysis doesn't suggest anything credibly. That doesn't mean anything is wrong with the four percent rule, but the truth of a conclusion doesn't dictate that every argument in favor of that conclusion is valid.

pecunia

  • Bristles
  • ***
  • Posts: 456
Re: trinity study in another dimension
« Reply #18 on: May 13, 2018, 09:50:33 AM »
Quote
As long as men continue to be able to innovate and not make wasteful destructions happen via war or environmental disasters, he sees no reason nice market returns won't continue indefinitely. However, as we all know, innovation does not yield results evenly, but in unpredictable spurts, with plenty of learning from failure in between.

One of the things that is mentioned a lot on the corporate news is "consumer confidence"

I've just been listening to a talk by the guy who made the Zeitgeist movies.  You know - These were the movies that said the whole thing was based on debt and we are all sitting on a stack of cards.  And - I think maybe it is.

Why does it work?  People want it to work.  They work to make it work until somebody changes the rules and it falls apart.  There's a lot of people in this world that have never had it so good.  Now they see movies and the internet and they see how they could have it good.  There's a lot of pent up human capital out there that want to make their lives better.  All of these people are going to be working for the corporations and starting new ones to make things better for themselves.  They will make things better for all of these other people that have never had it so good.  This currently wasted human capital will innovate, sell new products and develop entire new un-imagined industries.  Technology is not changing at a linear rate but at an exponential rate.

The stack of cards will be propped up,  the system based on debt will work for another generation or two and we'll be able to withdraw our 3.8-4 percent a year.

Just another narrative explanation / extrapolation, but a positive one.


Bicycle_B

  • Handlebar Stache
  • *****
  • Posts: 1868
  • Mustachian-ish in Live Music Capital of the World
Re: trinity study in another dimension
« Reply #19 on: May 13, 2018, 12:05:16 PM »
Since we're including very broad analyses in the discussion, let me comment on the house of cards thesis, and the up-and-coming-capitalists response. 

In my view the essence of the house of cards theory is that debt (or some other quirk of our particular system) is so critical to our society that once it cracks, the entire system fractures and then collapses.  Presumably the capital markets would then fail to deliver returns of 4% on a Trinity-style portfolio of stocks and bonds.  Put differently, adding debt has an accelerating effect temporarily, but later is fated to have a negative effect that will be stronger than the acceleration, so strong that it invalidates the "4% rule."

I suspect this is incorrect on two levels.  Returns may fall below 4%, but if they do, it will probably not be because debt was a fatal flaw, which if avoided would have allowed 4% returns to continue. 

The first level of incorrect is that debt is part of the surface expression of our society's wealth, and part of the mechanism for arranging our work, but it is not a fundamental part.  Developing new technology and maintaining old technology, growing food and maintaining tools, productive activities - these are wealth creation.  Debt, credit, stock, accounting - these are measurement tools, communication tools.  They are key parts of how we choose to arrange our labor, but the labor itself creates the wealth.  If all the debt in the world disappeared and everyone woke up the next day and grew food and drove trucks and handed food out to people who went to grocery stores, everyone would still be able to eat.  Sure, people are not going to change overnight the mindset of how we do things, so yes a debt collapse will temporarily cause a slowdown, but the fundamental economy underneath remains. 

The second level of incorrect would be that debt (or any other quirk similarly discussed) is so sneaky and two-edged a sword that when it goes bad instead of good, it will be so powerful as to destroy us, or at least the 4% rule. I suspect that our global economy will always have elements of instability, but then also develop responses that provide acceptable stability. What we survive, we will learn to handle. Since as discussed above we are going to survive, when debt (or whatever) periodically triggers a bad period, we will respond in ways that address that particular error.  If we fail, we'll learn the lesson twice and do better the second time. My guess is that the process won't be so bad as to invalidate the 4% rule...though it must be noted that the 4% rule only projects a 95% chance of being adequate for 30 years, not 100%.  Occasional dips under 4% are expected!

Re the up and coming capitalists - I agree they are a strong force sustaining future returns, probably strong enough to overcome debt headwinds.  Not strong enough to make the 4% rule stronger or weaker than before is my guess.
« Last Edit: May 13, 2018, 12:14:29 PM by Bicycle_B »

pecunia

  • Bristles
  • ***
  • Posts: 456
Re: trinity study in another dimension
« Reply #20 on: May 14, 2018, 06:36:22 AM »
Quote
The second level of incorrect would be that debt (or any other quirk similarly discussed) is so sneaky and two-edged a sword that when it goes bad instead of good, it will be so powerful as to destroy us, or at least the 4% rule.

Seems like there has been a lot of countries while not destroyed by debt, they have surely been crippled by it.  I guess Venezuela is one current example.  Greece has had it's problems.  Argentina used to be a very rich country.  Germany had sort of a debt after World War 1.  That exacerbated future problems.

China is doing well because they are the makers and sellers of stuff today.  The debt is owed to them 

The businesses in the countries where debt has hit them bad have not done well.  Hopefully, the folks in charge of the US have the competence to deal properly with debt and exercise the proper discipline.  However, lately, I've had my doubts.  Policies have led to short term gains in the stock market, with little longer term thinking.

I guess if it goes bad, I'll just shift from US to world index funds and get my 4% that way.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #21 on: May 14, 2018, 06:48:07 AM »
Quote
The second level of incorrect would be that debt (or any other quirk similarly discussed) is so sneaky and two-edged a sword that when it goes bad instead of good, it will be so powerful as to destroy us, or at least the 4% rule.

Seems like there has been a lot of countries while not destroyed by debt, they have surely been crippled by it.  I guess Venezuela is one current example.  Greece has had it's problems.  Argentina used to be a very rich country.  Germany had sort of a debt after World War 1.  That exacerbated future problems.

Debt denominated in someone else's currency certainly has the potential to be very destructive to a country. That distinction is important in the cases of Greece, Argentina, and Weimar Republic-Germany.

Venezuela's problems don't have anything in particular to do with debt, do they? It seems like a textbook example of Dutch Disease and/or the Resource Curse where long term high income from oil exports essentially caused the rest of the country's economy to atrophy, so now with oil not bringing in enough money anymore the country isn't self sufficient in basically anything, and doesn't export enough of anything (besides oil) to pay for the imports of food and medicine they depend on.

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1037
Re: trinity study in another dimension
« Reply #22 on: May 14, 2018, 07:21:52 AM »
The Weimar Republic was allowed to make reparations payments in steel, gold, coal, etc. The problem was the size of the debt, which was half a trillion of today's USD as adjusted for inflation. Another way to look at this debt was as 96,000 TONS of gold.

http://alphahistory.com/weimarrepublic/reparations/

The U.S. may have gotten itself into a similar predicament. Our national debt is spiraling upward (20 trillion and rising fast) and more and more of our taxes go toward interest payments rather than anything of tangible benefit to taxpayers. We're seeing the same result as was seen in the Weimar Republic: a loss of confidence in government institutions and democracy itself, a yearning for strong personalities who can deliver change for the working class by breaking the rules, minority-scapegoating, and nationalistic hubrus about "greatness".

Does it make any difference in today's world whether the multi-millionaires and billionaires receiving those interest payments are currently located in their Shanghi penthouse, their Carribean yacht, their Vancouver condo, or their Alabama golf course home? Of course not. What difference does physical location make? Accounts transcend nations. People see their taxes going into a black hole while their standard of living crumbles. That's why fascism is making a comeback. Why do you think Republicans are so eager to run up the debt? Debt shifts society toward the far right.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #23 on: May 14, 2018, 07:28:35 AM »
The Weimar Republic was allowed to make reparations payments in steel, gold, coal, etc. The problem was the size of the debt, which was half a trillion of today's USD as adjusted for inflation. Another way to look at this debt was as 96,000 TONS of gold.

http://alphahistory.com/weimarrepublic/reparations/
It's not a question of how they could make payments it's a question of how the debt was denominated, which in the case of the Weimar republic was in gold, which was a currency they didn't control. If you don't control your own currency, you cannot inflate your debt away when it becomes unsustainable and you can end trapped in a downward spiral.

Quote
The U.S. may have gotten itself into a similar predicament. Our national debt is spiraling upward (20 trillion and rising fast) and more and more of our taxes go toward interest payments rather than anything of tangible benefit to taxpayers. We're seeing the same result as was seen in the Weimar Republic: a loss of confidence in government institutions and democracy itself, a yearning for strong personalities who can deliver change for the working class by breaking the rules, minority-scapegoating, and nationalistic hubrus about "greatness".

Does it make any difference in today's world whether the multi-millionaires and billionaires receiving those interest payments are currently located in their Shanghi penthouse, their Carribean yacht, their Vancouver condo, or their Alabama golf course home? Of course not. What difference does physical location make? Accounts transcend nations.

It doesn't matter where the people who own the debt live. What makes a difference is whether those debts are denominated in dollars (which can be inflated away by the same government which owes the debt), or in some other currency.


ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1037
Re: trinity study in another dimension
« Reply #24 on: May 14, 2018, 08:40:04 AM »
I would argue that a politician running on a debt-reduction platform of engineering 10% inflation for 10 years could not possibly be elected. Would you donate to a campaign that promised to demolish your wealth?

Therefore, it is not realistically possible to inflate away the debt on purpose.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #25 on: May 14, 2018, 08:56:16 AM »
Well yes I would personally, because I'm intentionally not investing is bonds. But I agree most people wouldn't.

However your argument is predicated on the assumption that what politicians do after they are elected is determined by what they say they are going to do when they are running for office. I don't think that is a valid assumption.

pecunia

  • Bristles
  • ***
  • Posts: 456
Re: trinity study in another dimension
« Reply #26 on: May 14, 2018, 10:06:34 AM »
Quote
Would you donate to a campaign that promised to demolish your wealth?

Gentlemen:  I pose the question as to whether the value of the dollar is determined by elected officials.

Quote
Why do you think Republicans are so eager to run up the debt? Debt shifts society toward the far right.

It has been humorous in a black way that the GOP boys run up the debt and then complain when the donkeys suggest helping someone and they say, "OH NO, we've got too much debt for that."  I didn't hear much complaining of the debt being run up with the last tax bill.  I guess they didn't do much to help the average Joe.

How could the most advanced trinity type study ever track all of these human variables?  How could they predict what Mr. Trump is going to do?  Wars will affect the return on equities.

Bicycle_B

  • Handlebar Stache
  • *****
  • Posts: 1868
  • Mustachian-ish in Live Music Capital of the World
Re: trinity study in another dimension
« Reply #27 on: May 14, 2018, 10:35:37 AM »
I guess I should 'fess up - I feel kind of out on a limb with my debt comment.  Presumably too much debt can be a problem.  For example, it seems like a smart country ought to be cutting its national debt in good times so it can comfortably borrow in bad ones.  The US is harming itself long term by doing the opposite of that right now.  I assume that will result in a problem later.  I suspect that it will resolve itself within the bounds of historical normalcy that produced the "4% rule" data but can't prove it.

As a side note, I also feel like comments from the likes of Maizeman and Sol are a lot more valuable in this discussion than mine.  I learn a lot by reading here!

I partly jump into these discussions to ponder these things by writing about them.  Not sure I actually know what I'm talking about.  Don't take my remarks about the overall economy as any fountain of gospel wisdom. 

Here's an example of how I think about these things (wiser heads, laugh all you want, just point out the flaws so I don't have to repeat my mistakes).  Every few years, I read about the national debt.  It usually goes up (except when Bill Clinton is President), but its proportion of compared to GNP goes up and down.  Where is the change coming from?

Well, every year, the gov't spends more than it takes in, right?  Usually, yes.  Does it pay interest on prior debt, as has been pointed out?  Sure.  Yet the remaining debt also gets easier to pay because of inflation.  If it's $20 trillion today, inflation is 2% and the total rises to $20.4 trillion a year from today, the buying power of the debt is the same.  To me, it seems like the debt is only getting out of control when the combination of all three factors is more than the economy's growth rate, in the sense that up to that point, the debt could be paid without reducing anyone's lifestyle. Is that a crazy way to think?

I guess today is the day I double down on ignoring the old saying "better to be thought a fool, than open your mouth and remove all doubt".  Will await wiser comments.

PS.  Re "the Trinity study covering all these variables", it's not an automatic prediction machine that measures variables.  It's just an exploration of "what happens if the future is statistically similar to the past".  Nothing guarantees the future will actually be similar.  That said, to me it seems that politics and wars existed before the Trinity study and also existed during it, so their continued existence suggests roughly similar results in the future. 

If you want a scary proposition that could invalidate the Trinity study, use a model in which returns depend on economic growth, while economic growth depends on population growth multiplied by productivity growth.  World population growth is lower now than it was during the Trinity study, and is expected to fall further.  Is it plausible this could cause lower returns than the past?

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #28 on: May 14, 2018, 11:12:14 AM »
@Bicycle_B feel free to be self deprecating if you like, but I really do appreciate you kicking off this discussion.

Yeah, the idea that stock market returns depends to some extent on GDP growth, and GDP growth depends to some extent on population growth is indeed one of the genuine concerns I think people can bring up about trying to predict the future based on historical data.

However, it does need to be put into context. During much of the period that our historical stock market data is drawn from, population growth in the USA hovered around 1%. Today it's down to 0.7%. Even if we drop to zero growth, the flow through to stock market returns shouldn't be huge, although 1% at the margins isn't a trivial change either.

If we go negative population growth like a lot of western europe, things start to get scarier economically (even if that's probably a desirable outcome ecologically). One of the under reported correlates of the great recession is the the fertility rate in the USA dropped from 2.1 (just about replacement value) in 2007 to ~1.85 today. Without immigration the USA would already be well into negative population growth territory, although not as bad as Spain or Italy, which are already down to 1.3 children per woman.

ChpBstrd

  • Handlebar Stache
  • *****
  • Posts: 1037
Re: trinity study in another dimension
« Reply #29 on: May 14, 2018, 11:45:36 AM »
Makes sense, Bicycle_B. Bad news existed in the past too. Imagine being long equities during the Depression while the Nazis were rising to power. Who knew decades of growth lay ahead?

The turn of the conversation toward discussing the effect of the national debt is based on reasoning about why the 21st century will not be a repeat of the 20th. Our real question is whether the 4% rule, which worked well in the U.S. during most of the 20th century, will work well in the U.S. in the 21st century. What might have changed that would invalidate our assumption that the U.S. investments marketplace will continue to deliver sufficient returns to retire on a 4% WR? Well, the indebtedness of the U.S. - and the West in general - changed a lot.

As you pointed out, population growth also changed. I'll add a third: migration of economic growth to Asia, where some investment marketplaces are less developed and transparent, means that the returns on a worldwide portfolio held by a person in the U.S. may lag actual worldwide growth. Good luck holding shares in a communist country.

There are many economic upsides too. Fewer people smoke, we're better educated, crops are rotated, people wash their hands, IQ's are rising, vaccines are available, and infrastructure like railroads, harbors, and power lines already exist and only need maintenance. All this should result in people being able to produce more.

Lots of factors were unsustainable 100 years ago. Society sidestepped them with cultural, political, and economic changes. The suggestion was made that the U.S. could sidestep its debt through inflation (perhaps as a policy goal). However, creditors have a way of becoming the boss, and the high percentage of debt that is domestically owned just gives the U.S.'s creditors more leverage over the government.  When creditors, foreign or domestic, own an economy they can set it to liquidate mode.

maizeman

  • Magnum Stache
  • ******
  • Posts: 2548
Re: trinity study in another dimension
« Reply #30 on: May 14, 2018, 11:52:51 AM »
However, creditors have a way of becoming the boss, and the high percentage of debt that is domestically owned just gives the U.S.'s creditors more leverage over the government.  When creditors, foreign or domestic, own an economy they can set it to liquidate mode.

Are you aware of any precedents for this happening when the debt was denominated in the country's own currency, rather than a foreign currency or one backed by a commodity (like gold)? I don't dispute that it is possible, I just cannot think of any examples of this happening off the top of my head.

As for inflation as policy, in the long term our government faces choice between three options: dramatically cut spending, dramatically raise taxes, or do nothing (which will eventually result in either inflation or a default on the national debt, which would wipe out bond holders similar to inflation).

Based on my experience with our government since I came of age to vote, I'm comfortable placing my bet on "do nothing" but I acknowledge I could be wrong.

pecunia

  • Bristles
  • ***
  • Posts: 456
Re: trinity study in another dimension
« Reply #31 on: May 14, 2018, 05:06:08 PM »
Other than the quotes, here is more half thought out nonsense.

Quote
If you want a scary proposition that could invalidate the Trinity study, use a model in which returns depend on economic growth, while economic growth depends on population growth multiplied by productivity growth.  World population growth is lower now than it was during the Trinity study, and is expected to fall further.  Is it plausible this could cause lower returns than the past?

Other things are not stagnant.  Lots of smart guys out there creating a lot of new technology.  It has been pointed out that productivity per worker is high due to automation.  Lots of poor folks out there in the world who can become a good market for quality stuff.  And - Does an economy have to be based on ever larger quantities of stuff produced?  There's lots of stuff wearing out which can be replaced by better stuff.  Lots of infrastructure to be replaced.

I still think old Tommy Malthus might have been right as some world resources seem to be getting kind of stretched.  This is another thing that may not be fully accounted for in any past based statistical model like the Trinity study.  Maybe these stretched resources just represent another opportunity for innovation, a company will come along to help solve the problem and we'll get our 4 per cent.

India is soon to be the most populated country in the world - Looks like a huge market opportunity.  Population in Africa is growing.  Another market opportunity for someone.  Unfortunately, investment in those places will probably come from a country with a lot of spare capital rather than a lot of debt. 

Quote
As for inflation as policy, in the long term our government faces choice between three options: dramatically cut spending, dramatically raise taxes, or do nothing (which will eventually result in either inflation or a default on the national debt, which would wipe out bond holders similar to inflation).

If we have inflation and the dollar is not as solid as we'd like, another currency may soon dominate.  US may have to dance to their tune.  However, I still think we'll get our 4 percent.