Real Estate Numbers: Calculators and Spreadsheets Are Your Friend But They Have Limits

On a fairly regular basis I get email asking what I think of this or that loan calculator on the web, this or that predictive model for real estate prices or loan rates, etcetera.

Loan calculators are pretty simple when you get right down to it. Numbers go in, other numbers come out. It's just math - except that you've got to be careful about the numbers going in. Just because your balance is $400,000 now does not mean it'll be $400,000 after the refinance. It's very possible to do a zero cost refinance that adds nothing to your loan, but most people don't do it. Furthermore, I know I've said this before, but the only calculator out there that I trust is one that I know the provenance of. I've caught more than one company that had programmed its calculator to low-ball the payment. There's no way to tell for certain except using your own calculator, and if you have your own financial calculator, why are you using the web? You can cross check, however, because it's rare that two calculators will be mis-programmed to yield the same wrong answer. Also remember to add in closing costs and prepaid interest and escrow accounts, if you're going to have one, and always figure the cost of any points after everything else is added in there, because that's what the bank is going to do. Finally, don't take it for more than it's worth. Just because they tell you, "nothing out of your pocket," does not mean there are no closing costs. They exist. Somebody is paying them, somehow. Unless you know for a fact otherwise because you've discussed it and know where the money is coming from, I'm certain that "somebody" is you, and they're getting rolled into the balance of the new loan. I've had people bring me paperwork from other companies showing new loan balances thirty thousand dollars higher than they were expecting, with correspondingly higher payments. (I've also told people to never shop for a loan based upon payment more than a few times, also)

For spreadsheets, what you can get is usually an analysis of one variable per spreadsheet. I've programmed a loan comparison spreadsheet, but it only compares two alternatives at a time and it's not really suitable for use with the public, because you have to understand the limitations and GIGO factor. Just like I've got spreadsheets that answer the "rent or buy" question, among others, but you have to understand the limitations on the results imposed by your model.

As a computer programmer, I make a pretty decent loan officer. In order to compare financial information via spreadsheets, you have to understand what points of comparison the calculations are meant to compare. If your data is out of whack, if your assumptions are away from reality, or if you're trying to apply the comparison outside its design limits, what you get is useless.

I have several spreadsheets I have programmed and use. All of them have limits that need to be understood in order to get useful information out of them.

The first is a rent versus buy spreadsheet, that I first talked about in Should I Buy A Home? Part 3: Consequences. In that article, I spent a good paragraph telling you what my assumptions were in cranking the numbers. I think they are good and reasonable assumptions for the markets I have seen in my area in my lifetime, but many people might not. I just had someone make a comment to the effect that "rent doesn't increase with inflation." Well, it hadn't been keeping pace with the cost of buying the last few years when he said it, but that's not the case now. Even then, that wasn't the same thing as not increasing roughly with inflation. Furthermore, we've gone through a period when landlords were keeping rental rates low in the attempt to have someone else pay most of the mortgage of their investment property. Judging by the "loaf of bread" or hourly wage comparisons, or anything else except the price to buy, local rents have increased by a factor very close to general inflation over my adult lifetime. Whatever you think of my numbers, though, the fact remains that they are assumptions, and if they do not correspond to future numbers, the conclusions they reach have no bearing on the real world.

The second limitation upon this sheet is that it's assuming smooth increases. This is not what happens, as anyone over the age of ten ought to know. Over longer periods of time, the data may tend towards an aggregate average, but that says nothing about any given year. In reality, some years are plus thirty percent while other years are minus twenty. Even if my assumptions for averages are good, the spreadsheet that predicts the next thirty years is useful mainly to predict overall level of the market many years out. The numbers for any particular year are so much garbage, as far as the real world goes, where a 5% differential between estimate and actual is often enough to render something worse than useless. Even if my assumptions for average return are right on the money (and if I didn't think they were pretty close, I'd use others), any particular year could be at the top of a peak or the bottom of a market trough. If you know what state the market will be in in a particular year three decades out, why the heck aren't you richer than the ten richest billionaires in the world combined? Knowing what the market was going to do these past few years is a lot easier than knowing what it'll be like thirty years from now! I have what I think are good predictions based upon good models, but I don't have any god-level knowledge of where any part of the economy will be thirty years from now, and neither does anyone else. We see the future dimly, reflected through the present and the past.

Speaking of which, let's drag one of the standard disclaimers out and air the dirty laundry. "Past performance is not indicative of future results." Averages of past results may be the only way we have of predicting the future, but those results depend upon unknowable factors. Somebody could invent something tomorrow that utterly changes the face of housing thirty years out. You think the urban planners of the 1920s foresaw urban sprawl? I know for a fact that they didn't. What no model of the future can predict is unforeseen factors. I can't tell you what they will be or what effects they will have, but I can promise you there will be some. In 1894, Michaelson (who first measured the speed of light) said, "Our future discoveries must be looked for in the sixth decimal place." This just a few years after the formulation of Maxwell's equations, and within a year Rutherford had changed the atomic model forever, while the basis of quantum mechanics was being laid, and less than ten years later were Einstein and relativity. Michaelson was right in a technical sense that precise measurements were the key to unlocking future discoveries, but wrong in the sense he meant it, that all the major discoveries had already been made. My predictive model is more detailed than most, and I do my best to include all of the factors I see, but I have no way of including factors that I can't see, and one thing I can promise you is that there are some. It may work out that I guess right anyway, but that doesn't mean there weren't any unforeseen factors, just that I got lucky despite them. The further out the model goes, the more it is dependent upon subsequent events no one can predict. Someone could announce man-portable fusion power tomorrow, or "Star Trek" transporters, or any of dozens of new potential technologies that could alter the world, and that's just the technological possibilities. Politics and demographics will utterly change in the next thirty years (When I graduated high school, more people were predicting the world conquest of communism than the collapse of the communist system. Mr. Carter's presidency was not the United States' shining hour).

Just because we know that the precise numbers are wrong, however, doesn't mean that those numbers have no value in predicting the future. The way the numbers will move relative to each other is much more important information. Population is increasing and will continue to increase. Demand in major urban areas and desirable areas will continue to rise faster than supply, and since such areas are where most of us live or want to live, the price of real estate will quite likely continue to increase faster than inflation. Particularly types of housing which are universally desired, such as detached single family residences sitting on a certain amount of land owned basically fee simple. PUDs and townhomes are less desirable for most folks, true condominiums less desirable yet, and below that are apartments. Offer most people the chance to move up on the ladder of desirability, and they'll take it. Since the only thing preventing most people from doing so is price, price is what's going to make it ever harder to make that transition to more desirable housing. Living space in a desirable location is a scarce good. Living space, desirable location or not, is a limited good. The only way to change this is to somehow manufacture more space or arrange to have fewer people to share it. I'm not aware of any plans to manufacture enough space to make a difference to the billions of people on earth, so I'm guessing that barring worldwide nuclear or biological warfare, population density is going to increase, demand for housing is going to increase, and supply is going to stay pretty much right where it is. Nonetheless, this is only a guess. My guess is that housing will be about four to five times as expensive as it is today thirty years out. If it's only twice as expensive as today, we'll all still live in million dollar houses. If it's eight times, we'll be in four million dollar houses. The wider the net, the more probability I have of being right - and the less useful the information is. Unless the price right now is something like two cents, nobody sane is going to invest money for that long without a better idea of what the payoff will be.

Whether I'm right or not is something nobody knows right now, or even how close. Actually, not being quite that much of an egotist, the question in my mind is more akin to "how far off will I be?" But the data is still useful, because it tells me that as long as my assumptions are anything like real, we're all looking at living in million dollar real estate - the only question is exactly when. It tells me what people will be need to be able to pay every month, at least in a general sense, and it tells me that more and more people are going to get priced out of real estate, or down into less desirable housing, and that real estate is therefore going to be a quite satisfactory vehicle for creating personal wealth.

Each of the changes has consequences, assuming it does indeed happen. If it's difficult saving the money for the down payment when homes are $300,000, how difficult is it going to be when they're $1,000,000? At a guess, three times as hard. Fewer people will be able to do it, proportional to the numbers today.

On the other hand, no system of projecting the future is better than the limitations imposed upon it by limited foresight. If the population of the United States drops to 1789 levels all of a sudden - or 1607 levels - all bets are off. Of course if that happens, most of us won't be here to worry about it, and the ones that are will have bigger problems than the price of real estate. It's pointless to waste time worrying about the price of real estate in such possible circumstances, where the price of real estate would be the least of our worries.

Caveat Emptor

Original article here