A comment from a couple weeks ago has stuck in my mind, and having thought about it for a while, I wanted to give a more extensive answer. Here is the question:
Why did you wait so long to sound the tocsin?
Surely, you knew this 5 years ago.
I gave an immediate answer which you can read, in which I listed the things I was thinking about in relation to housing in recent years. But I think there is a deeper issue here and that is, what could we have known? What should we have known? I mean, just because it didn't occur to me (or Alan Greenspan) doesn't mean that the clues weren't there.
Now this is a dangerous exercise. It is very educational, however, to look back at your forecasts and trades to see what you missed and how you could have done better. When looking backward, one must be very careful to make an accurate assessment of what you could reasonably have concluded at some point in the past. It won't teach you anything to color your view of the past with knowledge you have now, but couldn't have had then. Human memories are notoriously inaccurate, particularly when colored by bias or outside influence. To wit, eyewitness (mis)identification is the leading cause of wrongful conviction in the U.S.
Anyway, since I haven't been blogging for the last 5 years, the only thing I can do is go back through notes, strategy memos, presentations, etc. to see what I thought in the past. Over the last few weeks, I've spent some time doing this, and here's where I've come out. I'd be interested to hear how the readers recall their thinking on housing over the same time period.
On Home Price Appreciation (HPA):
In 2003 I got a question from a client about rising home values. This client feared that home prices would rise to a point where people couldn't afford homes any longer. I argued that this was illogical. Its impossible for a price to rise to where no one can afford it. As soon as that happens, the price must necessarily fall. I think my point stands, but had I married this point with some points below, I might have been a bit more bearish on housing.
Several times in the past I made the argument that home price declines were unlikely, because secondary market supply of homes was highly dependent on positive HPA. Here is my logic: Joe Blo buys a house for $100,000 plus fees of $5,000. He puts 10% down. If at some point in the future he wants to buy a $200,000 house, he'll need $20,000 to put down plus say $8,000 for fees. Most people would have a hard time coming up with that kind of cash unless they have some capital gain on their house. So if Joe can't sell his house at a price that allows him to move up to the more expensive house, then he'll probably just stay in his current home. That means that if HPA is abnormally low, supply will decline, creating a floor around zero HPA.
And I think this logic would have held if it hadn't been for speculators. The theory I advanced above depends on people buying a house for housing. If someone is buying a house as an "investment" then they can become a seller at any price. There is no more natural decline in supply. I think the official statistics on homes bought for investment don't tell the story either. I've heard enough anecdotal evidence of speculators lying about their intent when applying for a mortgage to believe it was fairly prevalent. I don't know what percentage of loans were labeled as "occupant" which were actually "investment" but its some positive number.
So my view for a long time, probably dating back as far as 2002, was that home prices were likely to level out, posting a few years of zero or near zero growth. As each year passed and home prices continued to rise quickly, I extended the period of zero growth I expected. So by late 2006 I was thinking home prices would be basically flat for 5 years or so. But I would not have expected significantly negative HPA nationwide. A few metro areas, sure, but not nationwide.
If you combine the points above: prices needing to be at a level where buyers can afford it, and that speculators were going to add supply pressure to home prices, you have to conclude that HPA will be weak. I won't go so far as to say I should have known that HPA would turn negative, but it was much more likely than I thought 2-3 years ago.
Ultimately my view that HPA will be flat for 5 years might turn out to be right, but we might go through some actual negative HPA for a couple years followed by modestly positive figures thereafter. I think my supply limitation idea is valid, and its preventing HPA from falling more severely in the short-term. But with both speculators and foreclosures putting constant pressure on prices, its going to be a tough couple years.
On Subprime Lending:
I've been cautious on consumer loans for a long time, more than 5 years. Its been a while since I got involved in any credit card or auto loan ABS deals. I didn't have a specific fear, it was more than I didn't like the rapid growth of consumer debt and figured there was better soil to till elsewhere. Besides, as I showed in my post the other day, it isn't like ABS spreads were screaming out that I had to own 'em.
Starting in 2004, two additional fears rose in my mind. First was ARM resets. The ultra low rates and steep yield curve in 2001-2003 had cause ARM lending to grow rapidly. Anyone could see that short term interest rates were going to rise at some point, and many households would be faced with large jumps in their mortgage payments.
I thought there would be two impacts of ARM resets. There would be some defaults. But I thought the bigger impact would be reduced consumer spending. I thought that most households would be able to afford the reset, but obviously they'd have to spend less on something else.
Looking at the timing of resets, I thought this problem would be spread out over time. Here is a chart Bear Stearns used in a conference I attended in January:
Since the resets are spread out over time, I assumed the economic impact would also be spread out over time. Our current problems are less about resets and more about poor subprime underwriting, and we haven't seen much in the way of declining consumer spending. But looking at this table its fair to note that more resets are coming, and this could have an impact on consumer spending in coming quarters.
The other problem is mortgage equity withdrawal (MEW). My concern about MEW goes back to 2003. In 2005, MEW was adding between 1-2% to GDP, depending on how you estimate MEW's impact. Given that I expected HPA to slow to a crawl, it seemed clear that MEW would all but disappear. In my mind, that alone could cause a recession.
Like the resets, MEW isn't our current problem. Based on how consumer spending has held up, I'm not sure MEW is going to cause as much of a problem as I once assumed.
On Mortgage Lenders and Homebuilders:
I was bearish on both mortgage companies and homebuilder credit as far back as 2003, mainly because I thought they'd see their business decline when housing finally cooled. Normally when business in a certain industry cools, some company finds it expanded too fast and winds up in serious trouble, if not bankruptcy. Until recently, bonds for these types of companies weren't especially cheap, so it seemed there was little reward and plenty of risk. Again, I didn't think that loan losses or liquidity would be the problem, I thought it would be business volume. Wrong.
So my view was mildly bearish on housing, and more seriously bearish on consumer spending. For much of 2005 and 2006 I tried to play this by being long duration. That didn't do much either way, as rates were range bound for most of that period. Late in 2006 I got more neutral, as I grew more concerned about inflation. I put on steepener bets which have worked out well.
I also tried to play this in MBS, which I believed would start prepaying slower. The logic was that with lower HPA there would be less cash-out refis as well as slower turnover. Both those things are happening, but because MBS spreads have widened so much, this trade has been a miserable failure. In fact, the slow down in speeds is likely to continue, as now consumers of all types are having more trouble getting mortgage loans period. Still, trade isn't working.
I thought credit was going to be vulnerable as well. If economic growth slowed, I thought this would cause spreads to widen, especially given that lower-quality spreads were trading near historic tight levels. I had this view from 2005 on, and played it by owning higher quality credits and non-credit, high quality spread product, like taxable municipals and GNMA CMBS. This was a miserable failure as well, not because my view on credit was wrong, but because the liquidity crunch hit my off-the-run type sectors harder than it did lower quality bonds. This gets back to my post on liquidity crunch vs. credit crunch.
So despite correctly calling a housing slowdown, I did not foresee that the subprime problem would become as serious as it turned out to be. It wasn't until early 2007 that the prevalence of low-doc loans became apparent to me. I knew it was happening, but I didn't not understand the extent. Given that Alan Greenspan himself didn't realize the same, I shouldn't feel so bad. I think there wasn't anyone keeping stats on such a thing, at least nothing that was widely published and/or timely.
And even once I saw subprime was going to be a bigger problem, I didn't imagine the type of liquidity crunch that developed. I don't think I would have ever predicted that subprime losses would have such a large contagion in such a short period of time. That makes the fact that my two more prominent trades got hit so hard by the liquidity crunch tough to swallow. Its an event exogenous to the bonds I own. Its purely technical. And yet I'm getting killed.