The UCLA Anderson Forecast has deteriorated to market cheerleading and bottom calling. This year’s incorrect bottom call will not help their credibility.
Irvine Home Address … 12 Orangetip Irvine, CA 92604
Resale Home Price …… $494,900
I don’t mind the things that you say
I don’t even mind going out of my way
I try and do these things for you
Why should I do it
I’m always untrue
Well, I did you no wrong
Did you No Wrong — Sex Pistols
I attended the UCLA Anderson Forecast for Orange County last Thursday. The keynote speaker, Ramin Toloui an Executive Vice President for PIMCO, was very good. The speaker for the Orange County Outlook, Mark Schniepp the Director of the California Economic Forecast, was awful.
Commercial Real Estate Forecast
The forecast for commercial real estate was not very positive. The commercial real estate market is facing the same woes as residential, but with an 18-month lag. Rents are falling, vacancy is rising, financing is difficult to find, and most borrowers are over-leveraged. It will take many years for the commercial market to recover.
Ramin Toloui was an excellent speaker. He explained the solvency
problem of over-leveraged borrowers facing refinancing (he was speaking
about commercial, but the same applies to residential). A property
purchased in 2007 for $100M may have $80M worth of debt (it probably
has even more). This debt will need to be refinanced during the next 5
years. The value of the property has cut in half, and the new lender is
demanding 30% equity. When this property needs to be refinanced, the
borrower’s loan will be capped at 70% of $50M which is $36M; they need to roll over
$80M. The gap is too large to be overcome. If the spread were smaller,
creative financing may be able to bridge the divide, but as it stands,
we are going to see massive deflation in the commercial lending market.
The problem of insolvency Toloui described is the same facing ARM
reset debtors in the residential market. A property purchased in 2006
for $1,000,000 with little or no money down will be worth about
$800,000 when the ARM resets. A lender will look at comps and limit the
loan to 80% of 800,000. The borrower will need to come up with the cash
to finance the difference between $640,000 and whatever they owe. Not
many will have $300,000 sitting around, and many who do will not want
to waste it by dumping it into a depreciating asset. The FED is trying
to solve the problem of residential insolvency by lowering interest
rates. The commercial loan market will have no such luxury.
This presentation was the best part of the morning.
UCLA Anderson Forecast
According to the website of The UCLA Anderson Forecast,
“For fifty years, the UCLA Anderson Forecast has provided
forecasts for the economies of California and the United States. Founded by professor
Robert M. Williams in 1952, the national forecast has been recognized as one of the
most accurate, and has a reputation for being unbiased – a factor that the numerous
corporate and Wall Street forecasts cannot lay claim to. The UCLA Anderson Forecast
for California is the most widely followed and oft-cited in the state and was unique
in predicting both the seriousness of the early-1990s downturn, and the strength of
the state economy’s rebound since 1993.
I call bullshit.
What I saw on Thursday looked a bit like trained seals balancing a ball on its nose to get a feast of fish. Whatever objectivity and credibility they believe they have, it isn’t reflected in rigorous analysis leading to objective conclusions. Instead what is presented is a bit more like Gary Watts with a shotgun blast of statistics supporting a predetermined bullish(it) conclusion.
The main problem I have with forecasts like these is their lack of direct causation.
I have written before about Telling Good Analysis from Bad.
Once you have accurate data, the analysis of this data must focus on cause and effect. There must be direct causation linking a specific set of conditions to the outcomes these conditions will produce. A good analysis demonstrates this direct causal link in a clear and unambiguous manner. When an analysis relies on indirect causation, it is weak; when an analysis relies on implied causation, it is worthless.
In The Anatomy of a Credit Bubble, I demonstrated a number of direct causal links which impact how much people pay for houses:
- House prices are directly correlated with amounts borrowed.
- Amounts borrowed are directly correlated with the interest rate offered.
- Amounts borrowed are directly correlated with the borrowers debt-to-income ratio.
- Artificially low interest rates (reset issue) and exotic financing cause foreclosures.
- Foreclosures cause higher interest rates.
- Foreclosures above a certain threshold cause house prices to decline.
- Declining house prices causes more foreclosures. (note the causally related downward spiral)
- Declining house prices and increasing foreclosures cause lenders to lower debt-to-income ratios and raise interest rates.
- Lower debt-to-income ratios and rising interest rates cause amounts borrowed to decline.
- Less amounts borrowed (in conjunction with foreclosures) causes house prices to decline.
Notice the focus is always on correlation and causation forming a
chain of events leading to an inevitable conclusion. A good analysis
centers the debate around the premises. If the premises are true and
accurate, the conclusions cannot be denied.
In contrast, a bad analysis states a conclusion and offers support
through indirect or implied causation. When you read through the Gary Watts Real Estate Outlook 2007 you find yourself asking, “How does that impact house prices?” It is a question that is never answered.
The UCLA Anderson Forecast for Orange County is full of statistics just like a Gary Watts support, but the lack of direct causation weakens it significantly.
Residential Real Estate
I knew the presentation was in trouble when the speaker tells the audience to feel secure in buying a house because prices are at the bottom. I felt like I was being sold a used car or listening to a briefing by Baghdad Bob. He even called out the commenters on the blogs of the OC Register as “doom and gloomers.” I felt the camaraderie of the bubble blog world being challenged; besides, we were right.
The presentation is a series of charts and graphs similar to my analysis posts. There were a number of slides on defaults and foreclosures that looked very much like mine in Shadow Inventory Orange County.
There was a moment when the presenter was commenting on how defaults keep rising, but due to moratoriums foreclosures dropped for a time. I was thinking, “yes, that is shadow inventory.” But with a wave of his hands, he stops and says, “don’t worry about it, foreclosures will go down.”
Did I hear him properly? How can you lead people right up to the problem, show it to them, and then deny that it is there? He offered no explanation as to what happens to this inventory. He did say if there is any future inventory problem that it will be absorbed by rising prices.
What is supposed to lead us to believe that the UCLA Anderson Forecast is correct? Their say so? That plus a report full of fancy graphs and trivial statistics is all you get.
Other than perhaps agreeing with my conclusions or maybe John Burns (Webcast: US Housing – Recovery on Government Life Support?) who also says we have 15,000 units of shadow inventory in OC, what would have impressed me?
Timing the Bottom
Let’s say the UCLA boys had taken their wonderful data and applied some historic parallels and direct causation to call a bottom. That would have impressed me. The analysis might have looked like this (with some help from Calculated Risk):
- The last housing recession began in 1991.
- The recession ended in 1992.
- Unemployment peaked in 1993.
- Foreclosures peaked in 1996.
- The market bottomed in 1997.
Let’s look at the direct causation between these events and speculate on whether or not it should happen differently this time around.
The recession in the early 90s was caused by a slowdown in housing and real estate just like this one. That recession also saw slowdowns in defense contracting and other industries that made problems even worse. The recession ended in 1992, but the effect lingered as people had to be retrained to work in other fields, so unemployment did not peak until 1993. The delay between the end of the recession and the peak in unemployment is well documented.
There were many reasons for the foreclosure crisis of the mid 90s, and we have all of those problems back with more force. The foreclosures caused by unemployment do not occur on the day a borrower loses their job. The delay caused by draining all sources of savings, maxing out credit lines and utilizing legal maneuvers can slow the process for two or three years — as we have seen with properties profiled here daily; therefore, it is reasonable to assume foreclosures will peak two or three years after a major unemployment crisis. In fact, I would argue it is unreasonable to assume that foreclosures have peaked for this cycle — as the UCLA Anderson Forecast does — considering unemployment has not peaked, and the newly unemployed will cause defaults.
Last time around house prices bottomed as foreclosures peaked. It is unclear if either one caused the other. For example, if house prices bottomed simply because prices were affordable and supply was low, then foreclosures may peak not because borrowers are not distressed, but because distressed borrowers can sell into the resale market rather than go through foreclosure. Remember, foreclosures are not a sign of distress as much as they are a sign of distress that cannot be masked by selling in the open market.
The more commonly accepted conclusion is that once the pressure of distressed inventory was removed from the market — foreclosures ran their course — then prices rose because there was not overhanging supply keeping prices down. This explanation sounds reasonable, but is doesn’t explain why there was not a lag time between the peak of foreclosures and prices rising to work off the inventory. This lack of lag leads me to believe rising prices were partially responsible for falling foreclosures — something the UCLA Anderson Forecast is counting on this time.
Neither explanation of the coincidence in timing between the peak of foreclosures and the bottom of the market give us any indication of whether or not this phenomenon will repeat. I suspect it will not because the foreclosure volume is so large that there will be a significant period of time to work off the inventory, and contrary to the primary conclusion of the forecast, I do not think it is reasonable to assume that rising prices will magically absorb our shadow inventory because it is too large.
We will see who is right and who is wrong.
When will the housing market bottom?
I originally figured we would bottom in 2011. I was most recently quoted as saying I believe the bottom has been pushed back to 2012. Based on the facts and direct causations assembled above, when will the market bottom?
Well, we can throw out 2009 or 2010 because prices cannot bottom before unemployment peaks and foreclosures peak. On this basis alone, I am confident the UCLA Anderson Forecast is wrong. If unemployment peaks in 2010, and if there is a two or three year delay between the peak in unemployment and the peak in foreclosures caused by various delay tactics, then foreclosures should not peak until 2012 or 2013. If this corresponds to the bottom again, then we will bottom in 2012 or 2013. If we have a significant lag between the peak in foreclosures and the bottom of the market due to a glut of inventory, then we may not bottom until 2015.
Don’t do out and buy a house because you believe we are at the bottom. We aren’t.
Irvine Home Address … 12 Orangetip Irvine, CA 92604
Resale Home Price … $494,900
Income Requirement ……. $92,128
Downpayment Needed … $98,980
Home Purchase Price … $699,000
Home Purchase Date …. 4/17/2006
Net Gain (Loss) ………. $(233,794)
Percent Change ………. -29.2%
Annual Appreciation … -8.3%
Monthly Mortgage Payment … $2,150
Monthly Cash Outlays ………… $2,820
Monthly Cost of Ownership … $2,130
Baths 2 full 1 part baths
Size 1,689 sq ft
($293 / sq ft)
Lot Size 2,462 sq ft
Year Built 2005
Days on Market 8
Listing Updated 10/28/2009
MLS Number P708154
Property Type Single Family, Residential
HOME BUILT IN 2005 NEAR IRVINE HIGH SCHOOL. 3 BEDROOMS, 3 BATHS, BONUS LOFT/OFFICE WITH RECESS LIGHTING. MASTER SUITE HAS DOUBLE SINKS, SPA TUB AND WALK IN CLOSET. FORMAL AND CASUAL DINNING WITH FIREPLACE. LOW MONTHLY HOA THAT INCLUDES 2 POOLS, 2 TENNIS COURTS AND CLUB HOUSE. UPGRADED KITCHEN AND BATH WITH GRANITE COUNTER TOPS AND STAINLESS STEEL APPLIANCES.
These undesirable properties are getting pounded. This infill site is between the 5, a shopping center and an old condo development next to the high school. It has every combination of negative.