Today’s post is about how data, as powerful as it is, limits the dialogues you see for how to invest your money and, more importantly, what to do about it.
No one needs a post about how awesome data is. The most convincing argument for where to invest your money is to backtest a strategy and show how well it has performed historically. There is huge comfort in knowing that in the last 50 years, investing in the stock market has yielded a compound annual growth rate of ___%. It’s no guarantee of future performance, but it’s about the best information we have to rely on as to how the investment will do.
I agree that that historical performance data is still one of the most important, if not the most important, element to deciding on a strategy.
The point of this article, though, is to show you the link between data and what gets talked about by retirement experts, and highlight the holes where you may be leaving superior strategies on the table not because there is no data to back them up, but because the length of that data isn’t long enough to make it into academic circles.
Length of Data
Suppose you’re a bright-eyed, bushy-tailed academic. You’re going to change the world with a bulletproof retirement strategy. You sit down in the library, inhaling the sweet scent of books and good hard learnin’, and you get on your shiny laptop. Where to start?
Well, you know you need to back test any strategies you come up with for a suitable retirement period, maybe 30 years. You will want to have many data points to show in your set, which means you will want at least 50-60 years of data. What does that leave you with?
Cool, let’s get to it!
This is why you see a million articles on stocks and bonds. There are reports cut every possible way you can think of – 50-50 allocation of stocks and bonds, 90-10 allocation, rebalancing every year, every quarter, every second Tuesday of the month except on full moons…you know the drill.
Stocks and bonds have been around for 50+ years. They have enough data to measure across multiple time periods. So they get huge exposure in academic journals. The best studies get picked up in expert circles, which get picked up in media and, finally, by the average financial advisor.
What about newer stuff?
I’ll give you a great example. These are two different stock indexes. Which would you pick to invest your money?
I would pick the blue line all day. It has more years in which it is up, and when it is up it is up by a lot compared to the orange. Yes, there is also more downside, but you’ll notice that the performance in the good years more than makes up for the performance in the bad years. And the difference choosing between these two indexes can be as much as 10 whole percentage points or more! That would get your money to perform almost twice as well. Twice as big a stash for literally picking one market index to track over another.
So what is that blue line? Here are those charts again with labels.
The NASDAQ is heavily technology focused, which has more and more become our national export. It has not only beat the S&P 500, it has run circles around it. This kind of difference in return is unreal. You should be jumping up and down about a 1 percentage point difference in returns. A 10 percentage pt return? If you had a million dollar nest egg, it would be throwing off an extra $100k per year!
I am oversimplifying, of course. Certain strategies do better in certain environments and some folks might argue that there is technology bubble. However, one could equally argue that technology is now the main differentiator for the US going forward rather than manufacturing, agriculture, or consumer staples and thus will continue to outperform. That’s a longer discussion. My point with these charts is to show you how something really valuable might not even be on the table for discussion.
The S&P 500 created its first index in 1923. It arrived at its 500-company model in 1957. By contrast, the NASDAQ tracked its first set of companies in 1972. Any retirement researcher needing at least 50 years of data doesn’t even have the NASDAQ index on the table. They want to be able to show full 30-year periods and many of them in their report. Their reports trickle into the media and then to your local financial advisor. Don’t expect to hear about NASDAQ index funds as commonly as you will hear about S&P trackers. The Trinity study, which is widely cited in retirement circles, uses the S&P Index because the data exists for the period they wanted to study. But I’d argue 44 years of data is probably enough for you and I to want to start considering the NASDAQ even if the academics don’t.
There are a host of newer investment avenues that don’t have a long enough track record to have percolated widely through academic research papers. Some of them might be terrible, some of them might be amazing, but the point is that you can’t rely on them being raised to your awareness by academics and others given the pressures of their profession.
Being on the cutting edge is risky. Using smaller data sets is risky to one’s career as an academic. Among other ideas that fall into this category are peer to peer lending, alternative assets like private equity and venture capital, rental income via things like AirBnB, Real Estate Investment Trusts (REITs). You may see some work done on these topics, but they are not as well talked about, not necessarily because they are inferior, but because they are new.
Availability of Data
Even more frustrating than length of data is availability of data when considering our investments. You’re the smart, bright-eyed academic again. You want to change the world. You’d love to talk about peer to peer lending, or land speculation, or real estate. But that data is hard to come by.
So what do you do?
You move on, of course. You’ve got a job to do.
A few stubborn ones will try and mine or create their own data sets. These folks should be applauded. But this is hard going, and because of that, you cannot expect to see as much research done on those investment types.
The one I think gets the most short shrift which could really be a needle mover for you is real estate. Real estate has been around even longer than stocks and bonds as an investment class. But they are difficult to measure as an investment vehicle at the national scale. Stocks and bonds are traded on a national, highly liquid market. Your individual performance in real estate buying a house in King County, Seattle, WA is going to look very different than your Aunt Mabel’s performance buying a building in Pima County, Tucson, Arizona. How is a retirement researcher to capture all this in their research?
The best they can do is look at national averages. But who buys a basket of houses when they go shopping for a home? Is that what you asked for when you went to your local broker? With the advent of real estate investment trusts, you can get a national, accessible and liquid market for real estate but it is in my eyes a very different investment vehicle than buying local real estate because you lose many tax advantages, etc.
Real estate has several amazing qualities, not the least of which is the fact that it’s tangible and can be collateralized for a loan. Borrowing money to invest in stocks and bonds will not yield you nearly as good a loan structure. It also has flexibility – in a down market you can try and cash flow by renting a property out waiting for the market to turn right side up. If things are looking particularly hot, you can sell and just pocket the gains. You are in a local market which means you can probably get unique deals to maximize your returns – you are competing with fish your size rather than the high frequency traders and big beefy institutions trying to eke out advantages in the stock market.
So maybe you’re a little convinced. You must know several folks who have made their entire retirement goal from real estate like I do. In fact, real estate owners are some of the wealthiest folks I know and make up a huge chunk of the 1% in America.
“But what about historical performance?”, you ask. “Surely I shouldn’t just buy into real estate blindly.”
Of course not. The point was that this data at the localized level for decades is difficult for an academic to compile for a research project that can meet their professional needs – they may want to show a national perspective but real estate is hyper localized, or they can pick one market, but then their findings are not applicable to many others. But that data is findable by you. Local brokers have this data. You can check out Zillow and Redfin and Neighborhood Scout. You can cross-reference this with sample properties’ selling history to gut check that data. Academics don’t necessarily have the right incentives to research this avenue and publish the results which will trickle down through media into your ear, but that doesn’t make it not worthwhile. In fact, it could be a veritable gold mine of opportunity for you.
Length and accessibility of data influence what sorts of things experts research and thus what eventually trickles through media to our ears. Not all of the best strategies make it through these particular filters, and it’s your job to have a perspective on what best meets your needs. Foremost among the opportunities left on the table is real estate, which we will discuss in a different post.