 |
bostonbubble.com Boston Bubble - Boston Real Estate Analysis
|
SPONSORED LINKS
Advertise on Boston Bubble
Buyer brokers and motivated
sellers, reach potential buyers.
www.bostonbubble.com
YOUR AD HERE
|
|
DISCLAIMER: The information provided on this website and in the
associated forums comes with ABSOLUTELY NO WARRANTY, expressed
or implied. You assume all risk for your own use of the information
provided as the accuracy of the information is in no way guaranteed.
As always, cross check information that you would deem useful against
multiple, reliable, independent resources. The opinions expressed
belong to the individual authors and not necessarily to other parties.
|
View previous topic :: View next topic |
Author |
Message |
admin Site Admin
Joined: 14 Jul 2005 Posts: 1826 Location: Greater Boston
|
Posted: Thu Jun 03, 2010 12:57 pm GMT Post subject: |
|
|
GenXer wrote: | So, my guess will look like a uniform number in the following range: mean +/- sigma * sqrt(time). (emphasis added) |
You are using the very historical average that you are deriding others for!
GenXer wrote: |
Here's something to try at home. I may try it at some point. Take daily or monthly S&P500 prices - I have a set going back to 1950. Then calculate the 'mean' going back to 1890 (from Bodie's book), as a starting guess. And then simply try to see if 5 years out your guess matches the actual numbers, and see how far off you are 5 years down the road. Then shift 5 years ahead, and adjust your mean, and try 5 more years out, and see how you match. |
Again, this makes no sense for nominal stock prices. Using the P/E ratio would be far more appropriate and I suspect less chaotic.
- admin |
|
Back to top |
|
 |
GenXer
Joined: 20 Feb 2009 Posts: 703
|
Posted: Thu Jun 03, 2010 3:34 pm GMT Post subject: |
|
|
I never said I will use the mean other than the middle of the range of values, a HUGE range that includes the mean. It doesn't have to be mean, it can be the median. It doesn't matter! Its the range of values that matters. I can pick a mean of 0 for an idealized random walk. Its the huge possible range of equally valid outcomes that matter, not a single point that's included in that range (i.e. the mean, or the median).
And yes, using statistics blindly is bad, and that's my point.
Lets look at P/E. I can show you studies that prove conclusively that nobody can predict a company's earnings one year ahead accurately, and consistently. And of course, we all know that the price itself is quite chaotic. While the denominator is not that chaotic, it is nevertheless far from deterministic. Divide the two, and all of a sudden you have something that's far from deterministic as well.
I hope it doesn't surprise you that looking at annual PEs from 1891, the maximum standard deviation is 6 sigma (which is pretty large) larger than Case Schiller maximum standard deviation, so I would expect similarly larger errors which can easily be 100% or even more. Again, you can convince yourself of that by applying a simple 'guessing' algorithm as above. |
|
Back to top |
|
 |
john p
Joined: 10 Mar 2006 Posts: 1820
|
Posted: Thu Jun 03, 2010 3:49 pm GMT Post subject: |
|
|
As far as the regression to the mean type of thinking.... I can't believe I'm wading in to this (God help me)...
Anyway, you really need to have consistency with respect to the properties of the material to see the elasticity to the respective modulus.
If however, the deformation was caused by a transformation of the original material being acted upon, how can you expect it to react the same?
I think the metrics ought to focus on whether or not Boston is the same place as it was back when the "historical norms" readings were taken.
For example, water boils at 100 degrees (God I hope I'm right...), and say you add a little olive oil to the water or pour in a beer, that boiling point is going to change.
This is how I'd do the exercise:
Take historical wages of Boston relative to the house price at the time and the prior five years of wage growth percentage.
I'd look at what percentage of the household earnings it took to service the mortgage.
I'd then compare that percentage of yearly wage growth to the P/E ratio. If say, wages were creeping along, I'd expect that the P/E ratio would be much lower (kind of like how it is in the Midwest). I'd expect that if the prior five years were boom years, the P/E ratio will go up because people look at it like a "payback period" and they can arrive at after a few lean years (i.e. grow in to their mortgage). I'd look more at the relationship of how a bubble of wage growth stimulated the notion of "growing into your mortgage", sort of a cause and effect understanding to the properties. Also, take into account that Boston was a more blue collar town and only in the past couple of decades was it mostly white collar. Beyond that, other properties that changed in the consistency were that we had more rentals before that are now condos. Also, buying power of a single income earner was much greater years ago, and the earning bubble from the dual income trend certainly changed the nature of the housing market.
Throughout Boston's housing market history you have different political context. Before you go (here he goes again...), think about what happens when people extend beyond their means as they have done in the past. In the 70's we had Jimmy Carter, a Democrat, who typically have more loyalty to the debtors than the creditors. Carter's policy was to allow for inflation and when people's salaries eventually went up quickly, their long term debt as a percentage of their salary decreased. People back then were able to grow into their mortgage. The creditors, mostly Republicans, were pissed because the money owed to them was worth less. In the late 80's / early 90's we had Republicans and inflation was kept in check, and you can see that for most of the 90's house prices were flat.
Depending on the political philosophy you either get currency manipulation or a long flat slow period for correction.
Guys, I'm not a right wing nutcase. What I'm saying is, if we get a recoinage of the US Dollar and inflation, our exports and services will be cheaper globally and there is a chance we will get wage inflation. If this is the case, P/E ratios will go back to "historical norms" because the "E" will increase. If Paul Krugman gets his way, this is what we'll have. To him "Jobs, Jobs, Jobs" means Stimulus, Stimulus, Stimulus. If this is the case, I see Boston prices holding and perhaps going up nominally in some locations.
If we get a political philosophy that says go on a spending diet, we're in for a longer correction. If this is the case Boston home prices will stay flat and drop depending on the areas of job losses (desperate sellers). We also will see fewer people qualifying for mortgages because the mantra will be to go back to historical norms of 20% down payments and you'll get fewer qualified buyers which will put downward pressure on houses at certain price points and locations.
I'm saying look for "tell signs" as to which philosphy will prevail so you can plan accordingly. Because most people in Boston are Democrats, I push the Republican perspective only because it is their blind spot. In my own personal financial planning, I am evaluating both potential futures (which is really what I'm just trying to advocate). I mean you can not deny that the Republican Bill Weld totally transformed the housing market in Boston when he ended rent control. That government intervention was huge and to not understand that transformation of the property of the material is totally losing sight from reality.
Now everything I have just said is true, but ignores the structural concern. I think that the Unions and Democrats have won the time being, but I'd be careful of this one structural issue. Barack Obama just said yesterday that the Bush Administration believed in an "Ownership Society" where you were "on your own" when it came to paying for health care, paying for day care, paying for your house, etc. The structural problem isn't that philosophy, it is that we are in a global market right now and we are competing against nations that don't provide health care as a right, day care, and a house for everyone. The macro problem is that our standard of living is much higher than our competition's. Sure we can stimulate as much as we want, but if people keep spending their money on foreign products the borrowed stimulus money will just go overseas.
As complex as this all is, I see a mixture of some people getting a lost decade and other people getting stimulated and currencies getting manipulated. Remember when I said you have to look at the consistency of the properties of the materials to understand their behavior, well what we have going on today is foreign nations buying our debt, which dillutes our own economic antibiotics. What I mean is that elastic properties like higher interest rates as a reaction to debt, almost like the freezing of water isn't going to happen because foreign nations buying our debt is like them putting antifreeze in the water so it isn't going to behave the same. You guys are looking at 100 degrees thinking we're still water and I'm saying look at the antifreeze that's being pumped in to get a truer expectation of the new freezing point.
Bottom line, is that without all this foreing purchase of our debt, we'd have already had inflation, higher mortgage rates, and lower house prices. I think that global governments understand that irresponsible behavior needs to be curtailed so it won't move quickly to currency manipulation, but I think eventually it will have to. I think that they want to make the irresponsiblity be as painful as to leave an impression on those that were irresponsible, but will dillute our debt by dilluting our currency and we rebalance by dillution of currency and we get wage inflation. In order for this to work right, we do need major reform i.e. we need college tuitions to get back to reasonable prices, get the price of prescription drugs to reasonable levels, and get China and the rest of the world to get a basic global set of regulations for things like leverage limits and such. I think we need to go through a period of grand standing and finger pointing and a longer period where people can't ignore these structural issues, unfortunately, before we deal with them. In the meantime, expect protectionist measures, expect bailouts, stimulus (but don't plan on that for the long term). |
|
Back to top |
|
 |
admin Site Admin
Joined: 14 Jul 2005 Posts: 1826 Location: Greater Boston
|
Posted: Thu Jun 03, 2010 3:50 pm GMT Post subject: |
|
|
Alright... less abstractly... if we were to bet on the P/E ratio of the S&P 5 years from now, how exactly would you select the mean and sigma for your equation "mean +/- sigma * sqrt(time)" and what units of time would you use?
- admin |
|
Back to top |
|
 |
admin Site Admin
Joined: 14 Jul 2005 Posts: 1826 Location: Greater Boston
|
Posted: Thu Jun 03, 2010 3:55 pm GMT Post subject: |
|
|
john p wrote: |
Bottom line, is that without all this foreing purchase of our debt, we'd have already had inflation, higher mortgage rates, and lower house prices. |
That's definitely a major factor that has been holding off inflation, etc. A decline in the velocity of money has played an important role too.
- admin |
|
Back to top |
|
 |
john p
Joined: 10 Mar 2006 Posts: 1820
|
Posted: Thu Jun 03, 2010 4:29 pm GMT Post subject: |
|
|
Admin said:
Alright... less abstractly... if we were to bet on the P/E ratio of the S&P 5 years from now, how exactly would you select the mean and sigma for your equation "mean +/- sigma * sqrt(time)" and what units of time would you use?
I'd look at it as a percentage. I'd compare the capitalization of the S&P 500 to the P/E of the S&P 500 and then to the amount of money M3 in the Economy.
What I mean by this is when the babyboomers were in their peak earning years (during Clinton), they were shoveling decades of earned wealth into the stock market. You had more dollars investing in companies like these towers of capital sitting on these little sled dogs of earnings, which is why we had high P/E's. The total amount of money out there was enormous and this surcharge of babyboom nest egg wealth needed a home so it drove up the P/E's. Greenspan saw that stock market as being unstable for all that capital so he lowered lending rates to steer it into the banks using long term investments like real estate and eventually corporate debt. When you lowered the reserve rates he set off an even greater tower of capital that bloated the housing markets and sent the corporate world into acquisition mode, then came regulatory arbitrage where the money went to the place where the rules were lax.
I would look at how much money is out there M3, what is the capitalization of the S&P and evaluate how much of the capitalization is due to excess capital by evaluating the P/E ratio. What you will find is that we almost have too much capital but it never trickles down.
Another more scientific way to explain it. Imagine the M3 as the volume that the chemical reactions are taking place within. Then look at the states of the chemicals. The pressue acting upon the H20 will determine it's state i.e. vapor, liquid, or solid. The ever expanding M3 kept the H20 in a vapor state which prevented it from turning liquid, which we could have had trickle down economics where the rain drops could have fallen down. When Democrats pushed to lower lending standards, they increased the atmosphere of derivatives which kept trickle down economics from working. When George W. Bush saw that trickle down economics wasn't working due to the expansion of derivatives, his policy was to actually send checks to people, which is what I would have preferred Obama do. A big percentage of these checks ended up being spend on foreign products so it just accelerated the situation, but it did provide relief. My only big concern with Obama's approach is that it makes its way to the people and doesn't bloat government even further, which is a contributing factor to our debt. I also think we need to burn off some of this M3 and we do this by lowering leverage limits globally, meaning you can only borrow up to x times what you have on reserve. I think the EU needs to at least get into some for of Pay-as-you-go in order to keep one country from spending beyond their means.
And someone needs to invent an electric car that works great! |
|
Back to top |
|
 |
GenXer
Joined: 20 Feb 2009 Posts: 703
|
Posted: Thu Jun 03, 2010 5:02 pm GMT Post subject: |
|
|
admin wrote: | Alright... less abstractly... if we were to bet on the P/E ratio of the S&P 5 years from now, how exactly would you select the mean and sigma for your equation "mean +/- sigma * sqrt(time)" and what units of time would you use?
- admin |
The point is that we have a cone of uncertainty, so to speak, which grows with time. One way to define this cone is by taking your current mean, and your current sigma (assuming it is constant of course, which is not true), and assume a random walk-type scenario. Here's how you can do it:
http://homepage.mac.com/j.norstad/finance/risk-and-time.html
Scroll to the very end of this article, and this is how you can contruct a 'cone of uncertainty'. Norstad basically has a recipe for it in excel. You have to be careful with your units of time - it could be months, or years, but your sigma has to be consistent (i.e. if your data is monthly, use monthly sigma, etc).
The way I would do the betting game myself is as follows:
1) Take a large data set, annual Case Shiller prices for example, from Bodie's webpage
2) I start about 40 years out, just to have a 'stable' mean. Then I calculate the following (5 years out, annual data):
(data(time + 5) - mean(time)/mean(time) * 100
where time is running from year 40 to year, say 140 (appropriately scale years to run from 1 to the end).
This way I calculate % error in my guess, if my guess was the mean.
If you use the above to calculate the 'random walk' cone of uncertainty, then you can also pick a number uniformly in that range, and make that your guess. This is probably not a difficult project, but again, the results may not be statistically very significant. Either the mean or the cone of uncertainty guesses may be better, but it may not mean much for the future - the future results may differ. I'd also try maybe a 1 year guess -this way you'll get more data points.
I unfortunately don't have time to do this myself right now, but if you are going to try this, let me know - maybe I can play with this a bit to make it into a Monte Carlo type analysis. Also, this can be done with various indices, including Case-Schiller and PE for S&P. |
|
Back to top |
|
 |
admin Site Admin
Joined: 14 Jul 2005 Posts: 1826 Location: Greater Boston
|
Posted: Thu Jun 03, 2010 5:46 pm GMT Post subject: |
|
|
GenXer wrote: |
I unfortunately don't have time to do this myself right now, but if you are going to try this, let me know - maybe I can play with this a bit to make it into a Monte Carlo type analysis. Also, this can be done with various indices, including Case-Schiller and PE for S&P. |
I don't really have time either, unfortunately, and was hoping to get an idea of the ballpark for the numbers that would be used. My suspicion is that the range covered by a uniform probability distribution would either 1) include negative values for the P/E ratio (which wouldn't make sense) and/or 2) not extend far enough on the upside to cover P/E ratios which have actually occurred in the past (which would run counter to treating all reasonable possibilities equally). That suspicion is moot without numbers, though.
- admin |
|
Back to top |
|
 |
john p
Joined: 10 Mar 2006 Posts: 1820
|
Posted: Thu Jun 03, 2010 6:16 pm GMT Post subject: |
|
|
I didn't know "uncertainty" had uniform mathematical properties? |
|
Back to top |
|
 |
GenXer
Joined: 20 Feb 2009 Posts: 703
|
Posted: Mon Jun 07, 2010 12:13 pm GMT Post subject: |
|
|
john p wrote: | I didn't know "uncertainty" had uniform mathematical properties? |
This is 'make belief'. We are playing with idealized analysis - it's the only kind of analysis that one can actually do and get something resembling an answer (which is at least consistent with the assumptions - i.e. assume random walk, get an answer that works for random walk). In reality, risk depends on what's happening, that is, more volatility creates more risk, but there is always a hidden chance of instantaneous volatility (i.e. something crazy happening), so just because the markets act nice doesn't mean they can turn on you in a split second. So, I'd argue this would render our discussion with admin moot - our answer will not really work in the current markets with huge fluctuations. For that we'd need to consider the next step - using something that resembles a real market. In that case, markets behave sort of like random walk for a while, until they don't. People have been doing lots of math to deal with this, but so far nobody has anything that resembles an answer (i.e. no rule that is capable of allowing us to make any types of consistent predictions). |
|
Back to top |
|
 |
GenXer
Joined: 20 Feb 2009 Posts: 703
|
Posted: Mon Jun 07, 2010 12:27 pm GMT Post subject: |
|
|
admin wrote: | GenXer wrote: |
I unfortunately don't have time to do this myself right now, but if you are going to try this, let me know - maybe I can play with this a bit to make it into a Monte Carlo type analysis. Also, this can be done with various indices, including Case-Schiller and PE for S&P. |
I don't really have time either, unfortunately, and was hoping to get an idea of the ballpark for the numbers that would be used. My suspicion is that the range covered by a uniform probability distribution would either 1) include negative values for the P/E ratio (which wouldn't make sense) and/or 2) not extend far enough on the upside to cover P/E ratios which have actually occurred in the past (which would run counter to treating all reasonable possibilities equally). That suspicion is moot without numbers, though.
- admin |
I think you have a point. But it doesn't really matter at this point whether its PE or just a random sequence. I tried playing with numbers a bit, but this turned out to be a little more complicated than I initially thought. I'm not getting stable results - but I'm playing with 'generated' random sequences, not the S&P. Just trying to make this analysis robust (and potentially much more interesting). So in fact, if I do use the PE sequence, I'll have to 'limit' PEs to be greater than 0, you are right. The same way S&P index has to be above 0. So of course you can argue that you can't really have a number close to 0 for the S&P index. But who's to argue what the probabilities are for S&P being say 10 vs. S&P being 1500? The problem is, we don't really know what the chancese are of either, so a uniform distribution would be a very, very crude way of dealing with this. And you are absolutely right, some choices are probably more likely than others. We can assume a Gaussian distribution instead of the uniform - this will bias our guesses to be around the mean. I can probably think of a different distribution to make this even more interesting (i.e. my favorite power law). I'd still argue that simply due to the nature of the statistics, one has to guess a range, not a value, and picking a single value in that range will be error-prone, just as if you were to pick last year's mean. So when you say a value for the index is X in 3 months, you'd have to have exp(mu*t+/- sigma*sqrt(t)) error bars around it just to be more accurate (though not necessarily correct, as we've seen that you can be 5 sigma off easily)
But I think we are now getting deep into the woods here - the point is that there is just too much error and too many unknowns - even if we had this all coded up, we'd still be relying on bad assumptions. And like I said before, S&P is probably not enough data (and we'd have to generate synthetic data sets to make this a statistical test), but we'd still come up short because we don't have the generator that can give us 1000 insteances of S&P. |
|
Back to top |
|
 |
john p
Joined: 10 Mar 2006 Posts: 1820
|
Posted: Mon Jun 07, 2010 3:12 pm GMT Post subject: |
|
|
I think GenXer and Admin have this "top down" perspective where they are trying to observe and make sense mass amounts of data over time.
I don't have the mathematical background to do this.
I more look at it like a bunch of individual families making decisions that add up one way or another at any given time. I look at basic affordability, like what is the monthly income and average debt of a young family. Then, I try to look at their values to get a sense as to where they will invest their resources. For example, will young families feel comfortable with the urban school systems. Do they value "family time" and want to live close to home so they can spend time with the kids?
Given the current set of values and their corresponding demands (design loads), it makes sense that a place like Newton gets so much traffic, it embodies so many: proximity to work for time with family, it's a safe community broken down into villages so you get that smaller feel, the culture is a blend of small town New England with ethnic diversity and worldly perspectives etc. In fluid dynamics there is a term called "Static Head". It is like the pressure in a water main pipe and they say that 1 pound per square inch can raise water 2.3 feet or something. I look at it like how much people value Newton will be how high the price goes.
Now what if a family could afford a $3,250 mortgage and then all of a sudden, Greece's economy hits the skids and the Eurozone is rattled for a bit and everyone starts buying US Treasuries? Well, mortgage rates might drop as a result. And, then, you might see a little jump in house prices because the $3,250 mortgage can buy more principal at a 4.95% than a 5.75%. It's kind of like a block of ice can't flow through a sieve, but if you heat it up and turn it to liquid it can. I look for game changers that change the properties to allow for flow or solidification. An other example would be the "size of the pipe" that the pressure is flowing in. In Austin, you have tons of land so the pressure can flow outward, which keeps it lower. In Boston, because of the scarcity of land, pressure builds up more easily. One big cultural influence of Barack Obama to me wasn't just that he was African American, what he represented to me was a sophisticated family-oriented Urbanite. In Boston, we saw hipsters and double income no kids or retirees in the cities, but not so much those that wanted to raise a family, here people wanted the Norman Rockwell setting for the most part. In the past decade, this changed quite a bit. Because this was a recent trend, I often would press people to check under the hood to see if the "Good old Boy" graft network which made the City less attractive was still in force. If the City does in fact attract families that would be a total game changer for the Boston Market in my view.
I remember Lance Armstrong in this past Tour de France seeing a change in wind and then making a cut that shot him in front. I think seeing influences and making connections is part of experience and unfortunately younger buyers don't have all that yet. I'm wondering how we reach these guys without annoying them with high level mathematics or far out political or economic data? In the past, if you stayed within certain affordability parameters you had confidence that it would play out ok (which is why I ask you GenXer and Admin to offer what you think a decent set of affordability benchmarks is).
It's like you guys are looking for this Mandelbrot Set at a macro scale. I'm wondering if you see any consistency of behavior from macro to meso to micro.
Do you guys also think that economists understand the impact of how much influence an $8K tax credit is? I mean why was it $8k instead of $15K? That's the formula I'd like to have my hands on.
Lastly, I want to know what you think about the regionality of the market and where you think Boston fits in? Real Estate is pretty painful in Las Vegas and Miami and I don't understand how those markets ratchet down when we are using this uniform macro approach with government intervention. I mean why does Washington D.C. where people are paying over asking, get the same $8k deal as Florida? |
|
Back to top |
|
 |
CC Guest
|
Posted: Mon Jun 07, 2010 5:44 pm GMT Post subject: |
|
|
Quote: |
Do you guys also think that economists understand the impact of how much influence an $8K tax credit is? I mean why was it $8k instead of $15K? That's the formula I'd like to have my hands on. |
I thought about this question too. I guessed it may have something to do with national average house price and average downpayment. I don't remember the exact number, but US national average house price seems to be around $200k-210k. FHA is still giving loan to buyers with only as little as 3-5% downpayment (just like the bubble period!).
$8k tax credit means if you are buying an average house (or a cheap condo in a city like Boston), you basically only need a credit card to pay the downpayment and uncle SAM will pay your credit bill later through tax return.
$8k tax credit + FHA = No downpayment! |
|
Back to top |
|
 |
admin Site Admin
Joined: 14 Jul 2005 Posts: 1826 Location: Greater Boston
|
Posted: Mon Jun 07, 2010 5:48 pm GMT Post subject: |
|
|
CC wrote: |
$8k tax credit means if you are buying an average house (or a cheap condo in a city like Boston), you basically only need a credit card to pay the downpayment and uncle SAM will pay your credit bill later through tax return.
|
You don't even need a credit card. The Commonwealth of Massachusetts was itself giving out $8K loans for down payments, to be repaid with the federal credit.
- admin |
|
Back to top |
|
 |
CC Guest
|
Posted: Mon Jun 07, 2010 6:06 pm GMT Post subject: |
|
|
admin wrote: | CC wrote: |
$8k tax credit means if you are buying an average house (or a cheap condo in a city like Boston), you basically only need a credit card to pay the downpayment and uncle SAM will pay your credit bill later through tax return.
|
You don't even need a credit card. The Commonwealth of Massachusetts was itself giving out $8K loans for down payments, to be repaid with the federal credit.
- admin |
I didn't know that. Any link? The trick still works though.
======================
BTW, I should add:
$8k tax credit + FHA's ultra low down payment = No downpayment = prolonging the mess! |
|
Back to top |
|
 |
|
|
You can post new topics in this forum You can reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Forum posts are owned by the original posters.
Forum boards are Copyright 2005 - present, bostonbubble.com.
Privacy policy in effect.
Powered by phpBB © 2001, 2005 phpBB Group
|