The blogger Physician on Fire recently, and kindly, syndicated one of my posts about the historical ranking of asset class returns. I got an insightful comment there from another blogger by the name of Coach Carson, who focuses on
Two Kinds of Real Estate Investment
Coach Carson noted that the
We both agreed that the JST Study data, which I’ve used in numerous other posts, was one possible source of historical U.S.
Investing in the S&P 500 vs Rental Property
In past posts, I’ve considered the extent to which historical rental real estate returns might have exceeded stock and bond returns. But instead of using imputed rent data, I backed into this question by determining the level of net rents¹ needed to meet or exceed historical stock and bond returns. Working with data going back to 1963, I found that net annual rent payments of 4% would easily beat 10-year bond returns over the same period. However, net rents above 6% would be needed to beat the returns of the S&P 500.
Coach Carson’s comment made me realize that I could use the JST Study data as a cross-check on my previous analysis without necessarily fully accepting the accuracy of rents imputed all the way back to 1900. And conducting the same evaluation in two different ways is often a great way to verify conclusions or identify faulty assumptions.
Historical Rental Real Estate Returns
So, what does the JST Study say about the history of U.S. rental
- The S&P 500 – using Aswath Damodaran data
- Rental
real estate based on price changes only – using Shiller home price data. - Rental
real estate including rents – using JST Study data.
I started the dataset in 1928 because the further back you go, the less applicable the data are to today’s markets and economy.² According to the JST Study data, the nominal and inflation-adjusted annualized returns of rental
But that summary reduces the entire data set to investing over one very long period. Let’s look at how returns performance has varied over time by examining 10-year rolling CAGRs as shown in this graph.
I used 10-year rolling CAGRs because that’s a reasonably mindful investing duration. With that duration, there were times when the S&P 500 performed better than
The Real Estate Rental “Premium”
The rolling 10-year graph also shows a relatively consistent return “premium” provided by rents (orange line) over and above investing in non-rental
The 10-year annualized return premium provided by rents has varied from a low of near 2% to a high just above 7% over the last 93 years. The median premium was 4.83%. In other words, the JST Study is imputing a median annual net rent of roughly 5% since 1928.
Where Can You Get 5% Net Rent?
So, we would expect rental
Using the rule-of-thumb that 50% of gross rent typically goes to expenses, the median net rent across these cities as a percent of the purchase price was 3.78%. That’s quite a bit below the median 4.83% net rent imputed by the JST Study.
In fact, only 13 of the 78 U.S. cities had net rents above 4.83%. These high-premium cities ranked highest to lowest were: Detroit, Cleveland, Hartford, Memphis, Corpus Christie, Milwaukee, Lubbock, Baltimore, Panama City, St. Louis, Cinncinati, El Paso, and Aspen. On the other end of the spectrum, the 10 cities all with rents below 2.63% were: San Diego, Washington D.C., Salt Lake City, Irvine, New York City, Arlington, Seattle, San Jose, Honolulu, and San Francisco.
And out of personal interest, I collected data from Zillow for ten properties currently up for sale in my local town, which is a smallish city in central Washington State. I came up with dismal median net rents for my town of 2.59%.
If we plug in the median of 3.78% net rent from the U.S. city data and the 2.59% net from my town as alternative premiums for rental
The long-term nominal CAGR from investing equally in rentals across 78 cities in the U.S. would be almost 2% less than investing in the S&P 500. So, we’ve dropped another percentage point in returns from those assumed in the JST Study. And rental
Again, a more nuanced comparison is offered by looking at variations in the 10-year rolling nominal CAGRs since 1928.
I didn’t include the median from U.S. cities in this graph, because that line would simply lie somewhere between the JST Study data (orange line) and the extrapolation for my town (yellow line). Even assuming you’re investing in a town like mine, there are certainly times when rental
Conclusions
Historical data suggest that the lower your net rent, the less chance that your rental
However, I’m not saying that rental
The hitch with rental
My new takeaway from today’s post is that these data also show how great a diversifier rental
I’ve actually sold my two rental properties over the last few years because of the hassle factor. I slowly discovered that I didn’t really want to spend much of my retirement doing credit checks and repairing garbage disposals. However, after seeing the diversification potential of rental
1 – “Net rents” means the money the landlord (investor) receives free and clear of all related expenses such as mortgage payments, property taxes, maintenance costs, advertising, and vacancy periods. A popular rule of thumb is that gross annual rents should be near or above 12% of the purchase price for a rental to be a profitable investment. Another rule of thumb is that about 50% of rent typically goes to expenses, which equates to 6% in net annual rent as a goal for rental investors.
2 – Some of the reasons I’ve cited in past posts for why data from the late 1800s and early 1900s are less applicable to today’s markets include: prevalent corruption and market manipulation, only 1% of the population owned stocks as compared to 50% today, the economy was much more agrarian driven, and the Federal Reserve didn’t exist or function as it does today. For example, wild swings from huge inflation to huge deflation in one year were commonplace before the 1920s, and today’s Fed would almost certainly curb that sort of economic craziness.
I’m guessing leverage is the key to getting returns from real estate that outpace the S&P 500.
If you’re making these investments with 20% to 30% down, your realized returns will be significantly higher than price appreciation + rent, especially in low-interest-rate environments like we have today.
I enjoy having ample exposure to both asset classes.
Cheers!
PoF
That’s a good point that leads me to a couple of thoughts.
I was trying to address the question posed by Coach Carson on what the JST Study tells us about rental real estate returns. He thought that study indicated that U.S. stocks beat U.S. rental real estate over the last century or so, but not by much. And he was exactly right.
I’m not sure that comparing leveraged rental real estate to stocks is a fair comparison. After all, I can buy stocks on margin and boost those returns too, although the margin and interest rates available are better for real estate. I think that’s the better comparison because buying on margin (either real estate or stocks) increases the intrinsic risks considerably. And because those intrinsic risks are pretty different for the two types of investment, in my mind, it’s pretty critical to look at the balance of both potential returns and risks in such a comparison.
I’ve never bought rental real estate on margin because of my perception of the risks, which I discussed in this post. I’ve also never bought stocks on margin for similar reasons. So, I guess that also means I’m certainly no expert on the risks or returns on investing with borrowed money. Caveat emptor.
Thanks for the thoughtful comment.
When we want to diversify we want to look for assets that are mostly not correlated, especially when things are not looking good (i.e. crashes of one asset class).
Looking at your fantastic chart above, I can’t help but wonder, when stocks do crash, they usually indicate a very terrible economy timing. This is true for the last 3 crashes we had (2000, 2008 & 2020). And when economy goes down, wouldn’t it increase the risk of tenants not able to pay up rental and/or moving out? On aggregate, we could probably apply a % of probability to drop in rental yield to estimate the impact. But for individual rental RE investor, it’s either 1 or 0, it’s either the tenant continue paying, or you lose the rental income completely.
And that’s the problem with rental RE. You are just not diversify enough (unless of course you are mega RE investors with 50 houses). So if your RE’s rental ended up being impacted by economy, you drop right to the blue line in your chart and actually right in line with the equity performance at the worst of time. I understand that this line is 10 years rolling so maybe your underlying data will show otherwise.
Or maybe I am overthinking this as the JST study could already account for this. Just wonder what you think about this.
Thanks for the comment and yes, I agree. There may be less diversification (and/or lower returns) for individual rental real estate investors depending on how their particular rentals are impacted by the same economic turmoil that is impacting stocks. In general, I think it helps to separate volatility risk from intrinsic risk for any given asset. And I think you are talking about the risks that are intrinsic to rental real estate like renters who don’t pay, vacancy periods during hard times, large maintenance events that aren’t covered by insurance, etc. I agree that the chances of some of these intrinsic risks may go up in bad economic times.
Nevertheless, I think the 2008 period is interesting because it was a time when both stocks and real estate were hit pretty hard simultaneously. And yet, rental real estate (in aggregate) did quite well from 2005 through 2012. I’m sure individual landlords probably had huge issues in this period, but the JST Study data suggest that was not a common outcome for most landlords.
Really helpful analysis.
If you’re including rent payments, shouldn’t your also include dividend payments from the S&P, which will really boost the returns?
The S&P 500 returns include reinvested dividends.
STOP trying to compare stocks and real estate. No matter what metric you use they will never be comparable. I saw your previous post on PoF and agreed with Coach Carson. I did not comment then. The reason you can’t compare is because the real estate market is VERY INEFFICIENT. All your charts will NEVER capture all the returns an experienced investor gets from real estate investing. Let me give you an example, my last deal is valued at $210k today, it was bought 2 months ago at $165k, what chart or index tells you that I created 45k out of thin air? None. After buying I raised rent which was wayyy below market even without any renovation. What chart provides you this additional return? None. The tenant pays ALL their utilities and there is a manager so it is completely hands off. I paid 15% of the sale price and this tenant will pay off my mortgage in 15 years. Can I pay 15% for my 403b or IRA stocks and have someone else pay the remaining 85%? NO. I have a high income w2 job and pay a lot in taxes however I also have mid 5 figures of rental income annually that I don’t pay taxes on due to depreciation. Does any Index tell you that? No. I have tens of thousands in suspended passive losses from real estate that means I will likely never pay taxes when I sell the real estate or make money from any passive activity. Is this represented in the charts above? No.
Don’t get me wrong I have 60% right now in stocks and about 30% in real estate so I’m a big time stock investor so I’m not bashing stocks. Also I realize that real estate can be more work but if you don’t mind the little extra work you will get rewarded – unless you don’t know what you’re doing.
TLDR – real estate is inefficient. Lots of opportunities to make money and reduce taxes beyond Shiller or JST
As I said in the conclusion, this analysis reaffirmed for me that a balance of stocks and rental real estate appears to be an excellent diversified portfolio. So, just as your not trying to bash stocks, I’m certainly not trying to bash real estate. I’ve owned up to three rental properties simultaneously in the past. But where I live, I’ve never come close to getting 5% net rent or anything like that. On the other hand, I’ve done quite well in terms of price appreciation on all three of those properties when I decided to sell, which is similar to your experience. I fully agree that the range of potential returns from rental real estate is far wider and more idiosyncratic than the potential returns from stocks.
That said, I routinely compare the returns of many disparate asset classes, and so do others such as the JST Study itself. So, I don’t think I’m violating any laws of nature by at least attempting to understand the potential for returns between these two particular assets.
No. You are not violating any laws of nature. I’m just pointing out that trying to compare two things that are not comparable is an exercise in futility.
This is a late reply, but here goes.
I have had up to five rentals in the past. My net rental yields were around 5-6%, followed the 1% rule, 20-25% down, safe growing area, strong economy, etc. The difference between my purchases and your value-add example is that I bought mine turnkey, which didn’t require that much work.
How much time would it take an inexperienced person to find and do a deal that created 45k? To me that’s not passive, it’s a real estate BUSINESS. Which cannot be compared to index investing – maybe individual stocks, but real estate is probably much more consistent.
As for the real estate market being way more inefficient, I’d say that’s true – if you spend the time to take advantage of the inefficiencies. Syndications do this on a larger scale, and once in a while result in a total loss for investors, based on stats I’ve seen from Crowdstreet.
So I’d say you CAN compare real estate to stocks – turnkey rentals vs. index funds – this is more of an apples to apples comparison. The thing that annoyed me most about owning rentals is holding cash reserves – that’s cash sitting around not returning much – thus lost opportunity cost. Each $100k of mortgage debt I took on, I had at least $10k cash in reserves.
why not just having a LOC facility with a bank or two instead of keeping your own cash ?
New real estate investor here. Would love to hear your opinion.
Does your S&P 500 return include reinvested dividends? That would seem to be the logical correlate to including rents from rental properties.
Yes, I should have noted that these S&P 500 returns include dividends reinvested.
The main problem with rental real estate is it is a pain many times to deal with renters.
Interesting and informative. What about other indices such as nasdaq which don’t seem to give same returns.Do u invest in just trackers that follow s&p ?
I routinely use the S&P 500 as a measure of U.S. stock returns as do many others. The S&P 500 provides a reasonable approximation of diversified investing across a range of U.S. stocks. Also, researchers have developed a long data history for the S&P 500, which makes it particularly useful. There are certainly many other measures of stock performance that could be used in a similar comparison, but those comparisons would cover a shorter timeframe.
Personally, I do invest in S&P 500 index funds along with some other U.S. total market index funds and international broad market index funds. I describe my portfolio more here. However, none of that is an endorsement or advice for others without doing their own due diligence first.
You didn’t include 2 parts in your assessment
1. Return on Net rental equivalent of dividends, bc returns u have allude is w dividends reinvented
2. Tax advantage which is in ball park of 2.5-3.5% depending on which tax bracket you fall into ? It is not an easy sum.
Regarding item 1, I agree that one could assume that rental payments are plowed back into additional rentals. That would be similar to reinvested dividends. But that would be difficult to calculate and pretty specific for each investor. For example, let’s say you clear $500 one month on a rental property. You can’t invest that $500 immediately into another rental, because it’s not enough money to buy the next property. So, you’d have to wait to accumulate enough payments before you could reinvest. And how long you wait would depend on the cost of the next property, whether you are using leverage, and whether you have other funds you might be able to use to help buy the next property. Those difficulties are probably why the JST Study (the source of my rental property historical returns) did not attempt such a calculation, nor did I.
Regarding the second item, I agree that comparing tax burden for the two types of investment is also a difficult and investor-specific calculation that would be an entire post all by itself. And for stocks, it would depend on what kind of account (tax-advantaged vs. taxable for example) you are investing through.
Thanks for the thoughtful comments.