Hacker News new | past | comments | ask | show | jobs | submit login
The Rate of Return on Everything, 1870–2015 [pdf] (frbsf.org)
215 points by Sandman on Jan 5, 2018 | hide | past | web | favorite | 151 comments



Take a quick look at a graph of population growth from 1870-2015. It has an incredibly steep curve. Now think about how that graph will look from 2015-2100. Likely, it will be drastically flatter.

Wouldn't this have a huge effect on real estate -- indeed, isn't that population graph the primary driver of what happens in the real estate market?

I'm not sure the last century provides a meaningful guide here.


Yes, it will have a huge effect on real estate, but probably not the way you think it will. Real estate prices in desirable urban areas will continue to grow at the same, or higher pace. If you look at urban zip codes in desirable economic areas, they were barely affected by the 2008 downturn, and by 2010 the prices were recovered.

Today, those prices are much higher than they were in 2007.

Land supply is an almost straight line with a slight growth, probably something on the order of y = 1.2x, mostly due to innovations in transport speed, cars & metro. So unless we get maglev metro that travels at 200mph, RE prices will continue to reach new highs.


My bet is driverless cars, whenever they arrive, will actually increase the value of rural/small town land that's ~1 hour from a city center.

Hypothesis is many people live close to the cities because they hate commuting. But if commuting means work on a laptop or watch a movie, then the commute time is less significant. It's the near ring suburbs that will take the biggest hit. Small towns outside the suburbs will increase.


I suspect that commuting traffic will increase to eliminate this advantage. Already in major metros, long commutes suffer low speeds due to congestion during rush hour periods. Without a commensurately vast increase in infrastructure investment, I do not foresee a significant change in commute experiences from self-driving cars.

I predict that most people would not elect to live 2 hours away from offices despite being able to have access to entertainment and work while commuting.

The evidence is already here: various tech companies run shuttles to far-flung areas (Google has shuttle service from Stockton to Mountain View), yet employees aren’t moving to these lower-cost areas in large numbers.

Workers with families will have upper bounds to how long a commute they’ll endure. Their families are more important to them than being able to have entertainment, and meeting times limit the amount of time that can be used for work done while commuting. At that point, it’s approaching remote work, which is a promising idea, but isn’t a solution for long commutes by itself.

I see self-driving cars changing last-mile and replacing hub-and-spoke commutes as a replacement to taxi services and short-distance shuttles.


It'll work both ways, though. With ubiquitous driverless cars, offices may see less value in being in expensive urban centers and spread out.

Taken to its logical extreme: If people could teleport from one place to another instantly, there'd be almost no value at all to congested urban spaces. Similarly, if anyone can hail an affordable driverless Uber in seconds and reach their destination faster (due to improved routing / traffic flow and smaller cars) without having to think about parking, a bus schedule, or similar, the value in being physically close to other places decreases.

In that scenario, having your office out in the suburbs at a much lower cost becomes more reasonable. Transportation becoming more frictionless inherently makes longer distances more conceivable for travel (including commutes and office placement).


> Taken to its logical extreme: If people could teleport from one place to another instantly, there'd be almost no value at all to congested urban spaces.

So, like if we start being able to do our jobs in VR?


> I suspect that commuting traffic will increase to eliminate this advantage.

I think maybe in the short term, but in the long term I can imagine self driving traffic moving much faster than regular traffic. Cars all accelerating at the same time at lights, Uber pool in vans becoming super cheap. Roads could dynamically add and remove lanes in a direction without needing infrastructure, since all the cars will just "know" which lanes are available. Personal cars (hopefully rarer) can drive themselves home rather than take up parking space in cities.


The main advantage is the self-driving car could be an extension to your office. For instance if I have an 8 hour day with 4 hours of face-to-face meetings, then it doesn't particularly matter if the other 4 hours are spent commuting because I can work the whole time.


I don't think self-driving cars change the dynamics of traffic. Anything that makes driving faster attracts traffic until the advantage is lost. Often, it actually turns out to be worse due to second-order effects.


3D routes scale much better than 2d roads; also airplane autopilots are much easier to imolement.

I think self-flying electric vehicles will do what self-driving cars cannot.


I'm not so sure.

I've been taking an express bus into the city center from a rural location for a few years now. The time on the bus is between 35-50 minutes each way. Initially, I figured the same thing: I could get an extra hour of work done, work on side projects, etc.

The reality is that although the buses have WiFi, there are not many people working on them beyond the occasional email check. One major issue I found is that it's a lot harder to use a keyboard on something that's travelling on a road surface vs. a railroad. I even joked with someone about this last month as he was sitting next to me trying to write Javascript.

Most people on the bus are either playing video games or listening to music and I suspect that being that the people in this area are older and have families to get home to, having a more pleasant commute isn't remotely as important as having a shorter one would be.


that's a little American (and/or suburban) centric.

I and everyone around me live in urban centers because of the community and environment around your residence.

kids can just go out the font door and play. you get inspired by life and commerce by just looking out the window. etc

assuming you will drive and real state prices only impact how long is very californian.


A lot of people quite obviously love the suburbs. But the people who like cities want to live in the center. This means that as long as there exist some number of people who desire urban living, the prices there should continue to rise. What's being valued is the proximity to other people. And that supply is necessarily constrained.

Certainly some people live in city centers out of necessity. But the super-wealthy could live anywhere they want. They choose Nob Hill and SoHo (and so on). And they will continue to do so. Transportation options have very little to do with it.

Better transportation options for everyone else will be great! But the U.S. is an enormous place, so it's really only the city centers that should continue to see price pressures (because there's plenty of room everywhere else and better transportation options improve that situation).


Yes, some people prefer cities, but we have to ask whether those are the people fueling urban growth.

In San Francisco, I suspect it’s mostly jobs, not love for the city, that is causing growth and fueling housing price increases.


I live close to work because I like sleeping late and this allows me to sleep as late as possible while also getting in at a reasonable hour. A driverless car isn’t going I solve that.

The 15 minute walk/subway combo I have does solve that


Not only driverless but electric cars as well. House prices near the highways will increase significantly because the air pollution will become non-issue


Noise isn't a non-issue though.


Electric vehicles are much quieter though.


Having lived near highways and busy main roads, tire noise, horns and sirens are a lot of the mix.


Most have sound proof walls


If this was the case, could one observe such a trend already in places with proper public transport (light rail etc)?


Yes, there are still many areas with a lot of potential to get worse.

but there are certain areas that can't go much higher and sooner or later each area will reach a price point where prices can't go any higher. Ultimately, it just depends on how much people can afford to pay: which comes down to salaries, commute distance and # of wage earning people per roof. Sooner or later all the coping mechanisms will be exhausted.

The bottom line is, RE prices can't outpace salaries forever. If you double the number of wage earners per roof, you can double the RE price. But, after that it becomes harder and harder, as people aren't willing to live with multiple families per house.


Don't discount the impact of investors who have no intention of living in their owned real estate, though. For these individuals, there is no salary or commute distance factor to consider, so price sensitivity is very different. Thanks to globalization this is an increasingly large factor in rising real estate prices. See cities like Vancouver as an example.


Except, in order to profit, the investors have to eventually sell or rent the homes to actual residents, who would have to worry about salaries and commutes. Otherwise it's just a speculative bubble.


Do they? Lots of homes in places like BelAir sit empty, visited once or twice a year. Either as status symbols or a place to hide I'll gotten gains.


BelAir is also a desirable place to live--at least if you're rich--which means eventually the homes can be profitably sold to an actual resident. (And part-time residents are still residents.)


just wanted to add - the cost of debt and the desire to take on debt, and the laws around mortgage debt, also play a large role.


The recent decrease in the mortgage deduction will likely have some effect.


Unless it becomes more appropriate for more people to work remotely, which is the case. I imagine is one of the main reasons people move to urban cities is because of employment.


Employment, entertainment, education and culture.


And, quality services such as healthcare. Plenty of countryside with no good doctors or hospitals within an hour drive.


The suburbs have those as good or better than in the city though.


You're quite right about desirable urban areas. Example:

The first apartment I rented in midtown Manhattan was a 1BR condo in a fairly new building. The owner had bought it in 2006 for about $775k. In 2010, I looked it up on Zillow to see if I could afford to buy the place (I could not), and it was around $850k. Within a couple years of the crash, it had not only recovered, but appreciated a fair bit.

He sold it in 2014 for $1.2M, and according to Zillow it's now worth nearly $1.5M.


So, about doubled. Worth nothing that S&P500 also about doubled in that period (and quadrupled since the 2009 bottom), while paying dividends.


It's unlikely he purchased that property with cash though. And while you can use margin debt to achieve similar returns, you can't generally get margin called on a mortgage -- which I guess is the one main advantage.


The biggest factors in economic productivity are labor and technology (aka efficiency). Labor growth may be slowing, but technology growth is accelerating.


Besides your personal opinion, why would it be "drastically flatter"? Do you have a citation or any insight into the next 85 years that you could share?



The doomsday argument implies it will also be steep, but in the other direction, making real estate not so valuable.


I don't think anyone is planning their retirement savings based off a 95% chance of the only being 2 trillion to ever live - even at 10 billion living people you're looking at thousands of years for that prediction to be relevant.


I certainly hope so.

The biggest expense that the average person pays for is rent and home mortgages.

The world be auch better place if a >1/3 of everyone's paycheck wasn't going to rich landlords.


Why not move somewhere cheaper? You don't have to live in SF or NY or Seattle to work in tech.


Things work fine for me, because I am in tech. It is everyone else that I am worried about.

Housing costs are the biggest transfer of wealth in the world, and the higher costs are, the less money there is in the average person's pocket.


Highlights:

> 1. In terms of total returns, residential real estate and equities have shown very similar and high real total gains, on average about 7% per year. ... The observation that housing returns are similar to equity returns, yet considerably less volatile, is puzzling.

> 2. We find that the real safe asset return has been very volatile over the long-run, more so than one might expect, and oftentimes even more volatile than real risky returns.

> 3. ...our data uncover substantial swings in the risk premium at lower frequencies that sometimes endured for decades, and which far exceed the amplitudes of business-cycle swings.

> 4. Comparing returns to growth, or “r minus g” in Piketty’s notation, we uncover a striking finding. Even calculated from more granular asset price returns data, the same fact reported in Piketty (2014) holds true for more countries and more years, and more dramatically: namely “r >> g.” ...globally, and across most countries, the weighted rate of return on capital was twice as high as the growth rate in the past 150 years.


Some more highlights:

> In terms of total returns, residential real estate and equities have shown very similar and high real total gains, on average about 7% per year.

> The data summary in Table 3 and Figure 2 show that residential real estate, not equity, has been the best long-run investment over the course of modern history.

> Although returns on housing and equities are similar, the volatility of housing returns is substantially lower, as Table 3 shows. Returns on the two asset classes are in the same ballpark— around 7%—but the standard deviation of housing returns is substantially smaller than that of equities (10% for housing versus 22% for equities).

> Predictably, with thinner tails, the compounded return (using the geometric average) is vastly better for housing than for equities—6.6% for housing versus 4.6% for equities. This finding appears to contradict one of the basic assumptions of modern valuation models: higher risks should come with higher rewards.

Seems the way the real estate market works has caused a big drain both on economic growth (as investments have gone into real estate rather than more productive products), and on economic equality.

So a Georgist land value tax does seem like a pretty good idea.


Real estate returns have been juiced by government policies for decades. The interest deduction, government backing of mortgages, and implicit guarantees of mortgage backed securities. Aka the cost of buying a house to the consumer is lower than it should be on a risk adjusted basis.


The cost to the consumer of buying a house is probably higher than it would be without these incentives, actually, because it ends up causing more money to spill into real estate.

It's the same with college costs: every additional dollar of federally subsidized loans made available to the public adds a dollar to the cost of college, almost necessarily so.


You've mixed up two things here I think.

Returns vs Cost of buying a house.

The point is that the government subsidies make the profits on buying real estate high, which leads to high prices.

So both are true: "high cost of buying a house" "high returns on real estate"

Obviously they can't be true forever, the upper limit I guess is the ability of the lower classes to rebel divided by their tolerance of being priced out of their own land.


The grandparent was describing the long-term vs short-term effect of broad subsidies.

The short-term effect of a subsidy is a reduced cost to the consumer.

If subsidies are generally and persistently available, this reduced perceived cost increases demand, causing increased prices. The ratio of price increase to amount of subsidy (and thus whether the consumer or producer benefits from the subsidy) will depend breadth of the subsidy and on elasticity curves of supply and demand.

Then there's the increased demand because the subsidy increases ROI. further increasing prices of the fixed input (land/existing housing) to the good (living space).


Thank you. That is roughly what I meant to convey.

Mind you, because these subsidies are never really temporary, I suspect the immediate effect on prices is always to raise them. And perhaps even temporary subsidies cause price rises just by increasing demand, though the price rises might only be temporary.


I was... responding the last sentence of the parent.

All economics and human societies, really, are bubbles. Very, very long-lived bubbles. Obviously growth on Earth is ultimately limited, so yeah, these very long-lived bubbles cannot go on forever. The real trick is to predict when any one bubble will burst!


If you look at Figure 1 at http://voxeu.org/article/rate-return-everything, the advantages of housing returns predate those policies by decades, although it looks likely that the policies have stabilized the returns since they have existed.

Edit: The actual paper breaks down the data between all data and post-1950. The advantages of housing appear actually greater pre-1950, before those policies.


All true, but since I can't get the PDF to load maybe you can tell me, how were the costs of real estate included? Calculating yield for real estate is not as obvious to me.


One of the other comments had an archive.org link; there's a fair section on that.


Yeah "juiced". Versus a company that can not only write off all interest and fixes (unlike a person). But corporations can also depreciate the asset.

But it's the individual homeowner that's juiced. Lmfao.


Corporations are just an extra layer of taxes in the individuals who own it.


And thousands more write offs. I can't write off my lunch at 50% either for simply talking about "work".

Having owned 7 small businesses in part or in whole, you get away with a hell of a lot if you try. Having been an auditor, companies do so many illegal things they get away with, the ethical isn't even in the same category.


And the implied tax of maintaining the corporation.


I do like seeing confirmation of what is patently obvious. Real estate is The Problem, at least in economically healthy areas.

In economically vibrant cities real estate hyperinflation has soaked up all the gains and locked them up in non-productive assets, serving to enrich only banks and rentiers at the expense of individuals and more productive entrepreneurial activity. I've been whining about this for over a decade.


Yes. And a big part of the problem is people's perception of real estate. This idea, that real estate is doing "well" when prices go up, is completely flawed. When RE prices go up, it should be considered a bad thing, a sign that the market is sick: supply is not allowed to increase sufficiently to meet demand, causing endless negative externalities, needless suffering for everyone.


It's because real estate inflation is considered a profit, while most other kinds of price inflation are considered a cost.

The "logic" of that argument betrays some interesting biases.

Of course if you privilege an asset class it will appreciate in value. This has nothing to do with conventional mechanisms of supply and demand, and everything to do with systemic political bias that rewards the ownership of certain privileged asset classes, while forcing the rest of the population to pay economic tribute to those who own those assets.


Thats because real estate is an asset, whereas inflation is about the price of consumables.

Value increases of assets have never been considered inflation.


Could not agree with you more. London is a classic example at the moment! Even after 5 years of working in the city, most people struggle to buy a studio apartment in a bad neighborhood.


What is an "economically healthy area", and what is "more productive economic activity", if the return on capitol is much larger than the growth rate?


Economically healthy means there are jobs with healthy wages, etc. Los Angeles, San Francisco, and Boston would be healthier than say Detroit, Cleveland, or Bakersfield.

Productive means the money causes things to actually happen like new product development, entrepreneurship, research, health care, etc., vs. just sitting in a bank. Productive also means it benefits people doing real things vs. just collecting rent and sitting around.

Look at Silicon Valley and imagine how much more angel capital there might be if real estate were not so inflated. When someone buys a $1.5M starter home, that's $1M not available for other investments.


> When someone buys a $1.5M starter home, that's $1M not available for other investments.

Why do you feel the money disappears? From my perspective, the $1.5M is now in the hands of the seller, who is equally free to invest in other investments.


If you are buying and selling real estate you are de facto wealthy. Money in the hands of the rich is substantially less stimulative to the economy than money in the hands of the poor, because a dollar in a poor mans pocket goes towards real tangible demand for goods and services they want to spend it on, while money in the hands of the rich goes into investment accounts (stocks / loans) and money sinks (like real estate or futures) that are only meant to make more money.

This is the fundamental mechanism of why wealth inequality is bad for everyone. It slows the economy down if you are moving money from consumption into investment and never seeing it come back out (and income trends in the last 30 years / average wealth / income mean vs median all demonstrate that wealth goes to the top and stays there). Loaning that money back out just deflates the wealth of the poor even more, which increases debt, lowers net wealth, and the general availability of credit for small business creation is pretty damn low compared to decades ago, and the success rate of new businesses is awful. This means the money never goes back down the ladder - it gets swallowed up into money markets that move money around to make money while all the profits are just the productivity of the working class siphoned off.

So when working class people are giving larger and larger portions of their income back to the bankers that loaned them the money to buy a house from a real estate baron, that money is dramatically less stimulative going into the bankers investment accounts (be they stocks or <rarely> further loans) than if it were being put into raw consumption by the homeowner.


I think what you describe is true for the secondary market, but not the primary. Buying real estate and hoping market forces solely will help you generate a return would seem a poor use of money for the purpose of increasing the general quality of life for everyone.

On the other hand, real estate development, turning a barren lot into apartments with amenities, or single family homes, would seem to be productive and helpful, both are investments.

The same is true for investment accounts, secondary market trades in stocks and the like would match your analysis, but not so for IPOs, which presumably is being used for productive things by the company.


But market-based economics doesn't distinguish between those cases. If it's more immediately profitable to speculate, then speculation happens. If it's more immediately profitable to build apartments for the super-rich then apartments for the super-rich will be built. (And there will probably be a bubble in which the apartments remain empty "for investment", followed by a price crash, as numerous cities have discovered.)

What will not happen is affordable housing for the majority of the population.

Simplistic monetary analyses are incredibly bad at estimating social value, and even worse at maximising the returns from network effects that become possible when you allow for the benefits of wider wealth distribution.


Furthermore, if there is someone with money "just sitting in a bank" they

* they are idiots, because that is a poor use of their resources, and

* the bank is sending it out, so it is economically productive anyway.


> the bank is sending it out, so it is economically productive anyway

Even if the bank is just 'sending it out' to someone that is using it for property speculation?


'lending'

Sorry.


Until a recession hits and they are on a shopping spree.

There is a reason why hedge funds are slowly gaining cash positions lately.


Your definition of healthy wages is absurd. I work with a woman delighted to move to Phoenix from the Valley. She makes less, but has a house with a yard and constantly remarked on how cheap things are here. She’s living better despite her wages being “less healthy”.


That’s a really one dimensional analysis. Real estate has been and remains one of the major drivers of economic activity.

The stuff you are upset about is just the downside of increased productivity and consolidation. The only reason you have the wacky real estate costs in certain places is that the money people have concentrated in a few places. Their convenience is more meaningful than the inflated salaries required to support housing that’s 10x the national average.

People have a fundamental drive to improve their home to enhance its value but also be more pleasant and suited to their wants and desires. That drives everything from home improvements to furniture to art.

This ideal of a higher density residential society with lots of landlords that is popular on HN right now is a terrible future. Landlords want one thing — maximum return on assets. That means high cost, minimum possible opex.


> This ideal of a higher density residential society with lots of landlords that is popular on HN right now is a terrible future. Landlords want one thing — maximum return on assets. That means high cost, minimum possible opex.

How is that different from what homeowners want? Don't homeowners already block legislation that would allow higher density units to be built because it would decrease the value of their homes?

At least in the Seattle area where there are many high rise apartments and a lot of competition, companies seem to have to invest in making their apartments nice.


I fight against medium density development in my city because they are assessed by a different standard and my property tax burden is like 10x the developer. The development drives costs for school services and I get to pay.


> my property tax burden is like 10x the developer

Care to elaborate? Why would that be the case?


Valuation is based on net income, which is easy to game.

Also, most new development in my area is getting partial 10 year tax abatements. So the owners get to undercut the rents, which drives more of the traditional 2-4 family houses to lower rents and less maintenance, and eventually to section-8, which is the price floor.

So I get to pay more property tax, I get to pay for tax abatements via sales taxes, and we get to enjoy more traffic and a decline in the maintenance of legacy housing stock.


I'm not sure what the situation is for the OP, but in some areas, local governments give tax subsidies to developers. I like development, but I don't like tax subsidies for developers.


Density alone does not imply landlords, though. It is a very business centric view of housing to have density imply rent. In a simlar vein to how I wish more people would even talk about business cooperatives you can have a residential building that is also managed in a cooperative of its tenants. Such arrangements share swathes of the issues more general cooperative businesses have (like how nigh-impossible it is to start one) but there really are not only two roads out of the development hell we are in right now, where one leads to self determination where everyone is a property owner of a factory minted mcmansion with a standard slab of astroturf while the planet desertifies into an inhospitable fireball or everyone is subservient to tyrannical money hungry landlords in cramped conditions with rationed electricity and water access.


High density is interpreted by society as an inferior product, and there is no business model for coops at scale.

End of the day, only bigger players have access to capital without that federal subsidy.


Anything less than higher density in more areas is a catastrophic future for all of us. Sprawl is driving climate change. Sprawl hurts environmental air quality. Sprawl hurts health with long commutes and a lack of walking and exercise.


Sprawl only has a minor effect in all of those, and has many positive effects. My kids play in s 1/3 acre yard all day. for example.


Can they walk to visit friends, the local shops, park, school, library, etc? All of these things were much more important to me as a child than a big backyard.


I live in a 1800 ft^2 single family home on a 50x140 lot in a city. Our block is 50% single family, 50% 2-4 Family. We walk to school, library and grocery.

Because of the developer incentives, there is no incentive to build old neighborhoods like mine. 5-6 story sticks on bricks crappy apartments, or McMansions on cul de sacs rule the day.

When I looked at building a home, you can see why. It’s very difficult to build a 1700-2500 ft^2 home unless you go prefabricated, have cash, or can do the finish work yourself. The financing only works for real big or real small.


Yes, but usually ride their bikes. My town has 20+ miles of bike paths connecting the main parks/golf courses.


How does this all square with the Case-Schiller home price index, which showed something like a 1% real rate of return (i.e., just barely above inflation) for properties over time. As I understand it their data is based on comparable house sales and isn't subject to many of the common errors that occur when the mix of homes for sale shifts drastically.


Apologies for replying to my own comment, but apparently they are including the "imputed rent" in the calculation. That is, the amount of rent you would have paid if you didn't own the home. The thing is, you have to live somewhere, so if you didn't own the home you would have been paying rent. That's why the imputed rent isn't included in Case-Schiller, because as an investment you're not living in the home and it would be double counting to assume that you are. So I'm quite skeptical of this analysis that equities are the same rate of return as housing.


Can you link an article/journal on why a land value tax is good? I have seen this come up a few times in hacker news and yet I still can't find a good article/journal for why land value taxes are better than other alternatives and I have read dozens of economics/accounting/finance textbooks none of which back up land value taxes.


"In praise of the land value tax": http://theweek.com/articles/553242/praise-land-value-tax

The Wikipedia entry on the subject: https://en.wikipedia.org/wiki/Land_value_tax


Does 4 mean that the stock market is become larger and larger relative to GDP? That seems unusual. Can the ratio get arbitrarily large, or will it have to level off?


Yes, essentially -- this is a central idea of growing societal inequality, that economic growth gains are accruing much more to the rich (who own stocks and real estate) than to the working class (real wages not increasing).

Whether the ratio is now unusual, or how large it will get, is a subject of intense speculation with many real-world consequences for society, and especially to what extent law and government policies are responsible for it or can mitigate it (e.g. capital gains tax vs income tax).


What do you think pensions, IRAs and 401Ks are invested in?


You have a point but this is mitigated by the fact that most Americans have very little saved for retirement, just eyeballing the numbers it appears a tiny percentage of people are able to put away a substantial amount for retirement. http://time.com/money/4258451/retirement-savings-survey/


This article is very poorly written. First, how much you have saved should be directly related to your age, and they only touch on that in the second half of the article and only in way too big groups, ie not in any precise way.

Secondly, they don’t differentiate between married and unmarried. Two married 55 year olds with $300k each in 401ks are in better shape than individuals. As is someone with zero saved married to someone with $1M saved.

Also, they don’t include pensions and other defined benefit plans. That’s s substantial difference. If you are a school teacher two years away from getting $50k a year at retirement, you might not feel sane urgency to save.

and most importantly it ignores home equity. Paying off your home creates huge savings. I have a friend who paid off his $800k home by age 50, he would have answered that he has near zero retirement savings. He retired at 60 and built a million dollar home, no mortgage. Lots of people pay off homes by early retirement.

Finally, social security iften undercuts both the perceived need for big savings, and the means to do so. I’m paying in $9,000 per year right now, if i wasn’t i’d certainly contribute more to my 401k.


#4 means that the "rising tide lifts all boats" platitude is a lie. Or it means that the yachts are dumping their ballast into nearby dinghies.

I don't think the ratio can grow arbitrarily large, but I do think it can stretch the limits of reason for far longer that anyone thinks is possible before snapping.

The amount of the ratio below unity is the extent to which the ownership class captures gains from economic growth. The amount above unity is the extent to which the ownership class takes wealth from those with less wealth. That can theoretically rise as high as is necessary to make one person owner of the entire planet in one year. After that, the ratio drops back to unity, but by then no one cares to measure anything anymore.

If stocks rise 1.2 times growth, that means in aggregate, corporations captured 0.2 more value than was newly created that year, which means it had to be taken from existing wealth held by someone else. Some of that value might come from stockholders, and thus it would be recycled back to them, but the rest of it comes from non-stockholders, who lose it permanently.


I can't load the pdf but I'm surprised by the notion that real estate and equities have provided the same return over time. It's not what the Case-Shiller indices have shown.


i assume that they must be treating housing strictly as an investment, ie renting it out. In that case it sounds almost plausible: you get about 4.5% return in rent per year, plus an inflation matched housing value increase of roughly 2.5%, which gives you about 7% which matches the stock market.


The analysis is explicitly in terms of total returns, ie returns from change in price and from intermediate cashflows (dividends, rental income).

Edit to add: Total return = capital gain + yield


except, you don't get 4.5% return in rent per year, due to depreciation


I'm wondering what a realistic total real return is for purposes of investment/retirement planning. Seems to me that 7% is too aggressive, and one should work with about 4-5% or even less (given slower demographic and productivity growth).



I suspect that the issues with safe vs risky assets are a result of how screwed up the definition of economic risk is:

* Risk is defined by relatively short term visible price fluctuations, with "unusual" events removed from the model as "uncertainty".

* The risk model seems to use a Gaussian distribution, which doesn't fit the data particularly well.


I wasn't able to download the PDF but I read the summary article http://voxeu.org/article/rate-return-everything

In the paper, do they attempt to take into account the costs of real estate investment over holding stocks when calculating total returns? Things like property taxes, maintenance, stamp duties, realtor fees, legal costs dealing with problem tenants, etc.? If not I wonder if those explain some of the mystery around real estate returns relative to stocks?


Not to mention that real estate needs constant investment to keep it in working order and up to code. This is not the case with equities.


They use net returns (after operating costs etc., p.31), and provide an argument why leverage doesn't make a large difference (p. 40).


...and don't forget deal sourcing, tax prep - real estate investing is very labor intensive, vs stock market equities which is point/click and super easy.

For institutional investors, this is easily amortized but individual retail investors should invest in REITs to avoid over-concentration in a single property.


The PDF appears to be 2.9 MB and is taking a long time to download, possibly due to the source being "hugged" to death?

This appears to be an abstract, based on title: https://www.nber.org/papers/w24112

This paper answers fundamental questions that have preoccupied modern economic thought since the 18th century. What is the aggregate real rate of return in the economy? Is it higher than the growth rate of the economy and, if so, by how much? Is there a tendency for returns to fall in the long-run? Which particular assets have the highest long-run returns? We answer these questions on the basis of a new and comprehensive dataset for all major asset classes, including—for the first time—total returns to the largest, but oft ignored, component of household wealth, housing. The annual data on total returns for equity, housing, bonds, and bills cover 16 advanced economies from 1870 to 2015, and our new evidence reveals many new insights and puzzles.


There's a recent column by the same authors that's basically the executive summary: http://voxeu.org/article/rate-return-everything


Nice. The right side of graph 4, with the massive drop of r-g during WW1 and WW2, bolsters the argument made I believe by Branko Milanovic eg in Global Inequality, namely that war, by destroying fortunes, works to reduce inequality.



I wonder if there is a study of real estate investment returns in relation to demographics trends. For example, does poor demographical outlook (low births for instance) correlate with lower real estate ROIs later ?


I wonder if the extreme r - g divergence is helped by another factor: tax havens and hidden assets. GDP only includes things that can be counted and taxed; there have long been attempts to allow for the grey or black economy in GDP, but fundamentally it's dark matter that the owners are trying to hide.

But on the other end the assets are visible; we can count that there are X trillion dollars in bonds and Y trillion dollars in housing. We just can't work out who owns them because the trail goes dead in blind trusts in tax havens somewhere.

(There's a good paper on this that I can't find right now)


> GDP only includes things that can be counted and taxed

Unlike gross output, GDP only includes final outputs. All B2B activity is excluded by GDP, but capital is still involved. I wonder how things look if you look at gross output instead.


I was just thinking about this because Sears is in the process of closing something like 100 more stores. They were a behemoth of their time and now are barely staying a lot.

So many years of managerial incompetence and greed and market forces have completely eroded their dominance. I say eroded because they're not completely gone yet, and it's taken so long for them to glide down to this point.

It's just fascinating that company so large because reduced to so little over such a long span of time.


original link is dead (which could be the result of its sudden popularity on HN -- their entire server is quite slow for other content as well)...

archive.org has a mirror copy of this document,

https://web.archive.org/web/20180105120132/https://www.frbsf...


After taking a basic Mathematical Finance course, you'll find that nothing does better than the Prime Rate in the long term, unless you get lucky, or unlucky.


I just checked, and a portfolio of just the sp500 absolutely destroys the prime rate.

unless we are talking about two different things (what is the best thing to invest in to get the most returns over the long run), either you missed interpreted what the course was trying to teach you or its just factually wrong.

I took historic data from 1950 till now and matched it with the prime rate. I even gave you the benefit of the doubt by taking the max prime rate of the given year.

even if you invested at the peak of the dotcom bubble in 1999, you would have done better (100$=>255$ vs 246$)


Eh? I'd argue that the basic "theoretical" result (from equilibrium models such as CAPM and its modern cousins) is not that, but:

* for optimal portfolios: more risk -> more return

* however, idiosyncratic risk of a security doesn't matter (as it can be diversified away)

* what matters is systemic risk, ie correlation with the market return (or state of the economy): something that is negatively correlated (ie protects you by returning higher returns in a bad economy) is worth more, thus has lower returns


You may find it in that course but that doesn't mean you'll find it in real life...


its my understanding that its the sp500 that is very difficult to beat in the long run. Is that not correct?


The SP is not the 'risk free rate of return', but yes, it can be difficult to beat in the long run overall by individuals.


the risk free rate is not the best return. this is widely known. it is easy to beat the risk free rate, even if I take losses some years, I would destroy it in most others.


'I' does not mean all.

Individuals can, but overall, the net gains by all players is the risk free rate.


If you actually passed that course the professor should have their tenure revoked. The S&P500 rate is a rough proxy for the return on invested capital in the country, and it always crushes the risk fee rate over time.

I’ll give you a hint that may open your eyes. Volatility is not risk.


Better by what metric?


Overall Rate of Return over the long (infinite) term based on MC/random walk/binomial models.

It was quite a dissapointing result when I learned it.


People don't invest for an infinite term. In the long run, we're all dead.

Anyway models of low risk investments will underestimate long term crashes /Black swan.

Also, are you comparing other assets directly or somehow including combinations of them? Can you point me to a specific theorem you seem to be referring to?


True enough, we're all dead, but the mathematics can lead to a bit of depression when it comes to investing once you think about it a bit. Given all portfolio returns are based on the Prime Rate and modeled after random walks, it can be pretty easily seen that the rates of return based on modernish portfolio theory[1] will approach the Risk-Free (Prime) rate.

I wish i could find the actual theorem, but it's been a few years since I took the course and have long lost the textbook[1]. It was pretty clear though when it was taught, or at least that's what I gleaned out of that lesson without the teacher having to say it.

[1] https://en.wikipedia.org/wiki/Capital_asset_pricing_model


From your link:

E(R_m)-R_f is sometimes known as the market premium (the difference between the expected market rate of return and the risk-free rate of return).


Is the CAPM model an accurate description of reality?



There's a lot of criticism levelled at the CAPM, including significant methodological challenges (measuring the market portfolio, etc.), but the basic intuition underlying it is sound.


It's what is taught, but I wouldn't say it's entirely accurate though unless the market is fully 'efficient', which we know it's not, or at least not all the time.


It's provably wrong.

The risk free rate of return is reduced because people want to leverage short term cash flows. A 99% chance of gaining 5% is not necessarily worth a 1% chance of losing 5%. EX: Collage tuition is paid before teachers salary's are paid, so collages want somewhere to stuff money for a weeks, but losing money is vastly worse than some minor gains.

If you model the stock market by say buying evenly from all stocks and selling in 50 years repeat. Then some outliers like dell at IPO going up 500x more than makes up for losses. But, you can still lose a lot of money over say 5 or even 20 years and people can't necessarily wait 50 years.


All of those models depends on the specific axioms used to create them. It is not clear if the axioms are correct. The result is disappointing, but real world experience suggest that the result is wrong.


I guess nobody here has taken that course?


I think you need to elaborate more on what exactly you're talking about.

Nothing does better than the prime rate? You mean in terms of interest rate? Isn't the definition of the prime rate the best you can get?

In the long term? What does that even mean here? Can I get lucky and do better than the prime rate, whose definition is the best interest rate you can get?

Saying this knowledge is basic Mathematical Finance is also very condescending. I haven't taken "that course" but I feel from your tone that you took a 101 college course which now guides your entire mode of economic thinking.

Elaborate on your statements.


4th year Introduction to Mathematical Finance, but basically 101 in a sense.

Using a binomial model (essentially a random walk) and using 'the formulas' for portfolio calculation, you will find that in the 'long term' (in math, infinite, but 'a long time' in the real world') you will find that no investment portfolio will beat the Prime Rate, as this is what all returns are based off of.

It was just the results I saw from the course. I know one can get 'lucky' and beat the Rate, but overall, as a zero sum game, someone else has to lose and overall the group rate of return is calculated as...the Prime Rate.

Sorry for being condescending sounding; I thought this was common knowledege given it's taught in introductory finance.


Yeah I don't get why you'd assume 4th year Mathematical Finance would be common knowledge. It still strikes me as knowingly condescending.

I don't agree that someone else has to lose in order to gain. That assumes zero growth, which is not the case.

I think people don't know what you're talking about because you seem to just be repeating stuff you heard in a lecture hall 10 years ago when the world is vastly more complicated.


The stuff taught in introductory finance is a considerably simplified model of the real world. The prime rate is I believe the rate banks lend to large corporations and those corporations wouldn't bother borrowing unless they could get a return exceeding those costs.


To price derivatives one uses risk-neutral probabilities, which are different from real-world probabilities. The binomial tree you talk about cannot be used to calculate the expected return. Maybe this was not properly explained in the course you took, or maybe you were not paying attention.


It's really not.

Judging from your mentioning MC and binomial walks, you might be looking at derivative pricing, and be confused by the facts that

- derivatives are zero-net-supply securities, thus they returns are necessarily zero-sum (minus the exchange/bank's cut)

- derivatives can be priced using the risk neutral measure (after a change of measure, Girsanov's theorem, yada yada yada) in which mu, the drift of all risky assets, equals r, the risk-free rate. However, that's not statement about how mu is in the real world, but basically an abbreviation for an arbitrage-by-replication argument.

At any rate, derivatives pricing is the domain of arbitrage pricing models, while here we are looking at equilibrium models (which consider investors' preferences/utility, unlike arb models).


The linked paper is empirical counterevidence to that theoretical model though.

Admittedly, we can always just say "we're not in the long run," I'm not sure how useful that is.

I'm not certain what model exactly you're talking about, but I think it also probably misses technological change as a real source of growth independent of any monetary musings.


Theory is an 'efficient' market model. Practice is not. Reminds me of the old joke:

An Efficient Market Theorist sees a $100 lying on the ground, and passes it by saying "If that existed, someone else would have picked it up by now!".


> An Efficient Market Theorist

Milton Friedman, in the apocryphal story. And back in my days it was $20 :-)


The paper's data set is from 1870 to 2015, which is 145 years. If that's not close enough to "the long run" for the theory to match reality, then the theory is either wrong or useless.


the sp500 is not zero sum. you are ignoring dividends, and I think stock buy backs throws a monkey wrench into that assumption as well. lastly, population and economy increases.


Haven't taken a mathematical finance course... What is the long-term historical value of the prime rate?

I've seen data showing the overall stock market returns 3% after taxes and inflation for any period longer than about 20 years.


This is an empirical paper not a theoretical discussion on mathematically maximizing the formula for returns...




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: