• Commercial real estate prices are falling faster than home prices and losses will mount
• Private commercial construction will continue to contract sharply, hitting GDP growth
• Lenders’ losses may total $250-500bn, hurting especially regional and local banks
• Banks’ commercial real estate losses will keep them cautious on lending through 2010
• But CRE losses will only be 10-20% of total US financial system losses - housing is still the main story

Concern is mounting that losses in the commercial real estate market (CRE) are the next shoe to drop? The loss cycle is running behind residential mortgages and will eventually total to large numbers. But the scale of likely losses is much smaller than for the housing sector and we expect CRE losses to account for only 10-20% of total financial sector losses, about the same as for unsecured consumer loans. Nevertheless private commercial construction will be extremely weak for an extended period, dragging on economic growth and offsetting the likely pick-up in residential investment. Also the CRE bust hits particularly the smaller and regional banks and will keep them very cautious in their lending as the losses work through.

Sharp fall in commercial real estate prices

Commercial real estate (CRE) prices peaked in October 2007, about 16 months after residential real estate. Prices are now down 24% according to the Federal Reserve, though other analysts believe the fall is already greater. The Moody’s Real commercial property price index shows prices are down 35% from the October 2007 peak including a fall of 7.6% in May, the latest month. This means CRE prices have already fallen as much as house prices, measured by the S&P Case-Shiller index, even though CRE prices peaked over a year later. Another index, the NCREIF returns index from the National Council of Real Estate Investment Fiduciaries, shows total returns off only about 15% in this downturn. But if we allow for rents, prices are down about 25% which is already more than the decline seen during the early 1990s bust which followed the Savings and Loans debacle. Moreover the decline this time has been much sharper probably reflecting the very rapid tightening of credit terms. Another difference this time is that underlying consumer price inflation is much lower than in the 1990s and is not providing the cushion for nominal real estate prices that it provided then. For example in the 3 years from 1990 - 1992 core consumer prices rose a cumulative 14%, whereas we forecast core consumer prices from 2008-10 to rise only about 4%. It is difficult to be precise about prices partly because sales transactions have plummeted, from a quarterly rate of $195 bn at the peak to only $20 bn in Q1 2009. Of course most real estate is owned on a leveraged basis so investors’ equity losses are correspondingly larger. The MSCI REIT index, an index of real estate investment trust prices, is down 65% currently from the 2007 peak, despite moving up since March in line with the general stock market.

Prices set to fall further
The case for further price declines is compelling given rising vacancy rates, declining rents and the severe distress in the sector which is leading to foreclosures and forced sales. Moreover, financing for purchases is now only available on much tighter terms, if at all. The Federal Reserve’s Senior Loan Officer lending survey shows 66% of banks tightened loan terms in Q1. Meanwhile the commercial mortgage backed securities market (CMBS) is still effectively closed. Outstanding loans held by commercial banks are declining at present. One forecast of vacancy rates, from Marcus and Millichap, a large national commercial real estate brokerage, expects office vacancies to reach 17.6% this year (versus 12.6% two years ago), retail vacancies at 11%, (7.2%), Industrial at 12.6% (9.4%) and apartment buildings 8.2% (5.7%). Another segment suffering badly is hotels where in mid-July the occupancy rate was down 13% from 2007 levels. Room rates are also lower so the revenue per available room (RevPAR) is off 17.5% compared with mid-July 2008.

Falling commercial construction will drag on the economy

The CRE construction cycle normally lags the rest of the economy as projects begun earlier are completed in the early part of a downturn while, after the recession, high vacancy rates keep new construction slow for a while. This cycle has been no different. In 2003-6 non-residential construction bumped along at a low level in relation to GDP, only starting to rise strongly during 2006. But then it continued to build up right through to late 2008, even while residential construction spending was in a long decline. As recently as Q3 2008, commercial construction activity was still making a positive contribution to GDP (adding 0.36% to growth). But construction fell back in Q4-08 and slumped sharply in Q1-09, subtracting 2.05% from GDP (SAAR). Given the severity of the recession and the problems with financing we expect very few new projects to start over the next year (other than government schemes) which means further falls in commercial construction, probably taking construction spending below 2003 levels. The architectural billings index, a leading indicator of new projects 6-12 months ahead, also points that way. If we are correct that US growth over the next few years will be relatively sluggish (after an initial inventory-led bounce), vacancy rates will remain high and construction of new buildings will take many years to pick up again. Note that residential construction is following a very different cycle. We expect residential construction to pick up gradually over the next few years and be a significant contributor to growth as house-building returns to more normal levels.

Did the US overbuild?
The conventional view at the beginning of the recession in 2008 was that, overall, the US had not seen a huge over-build in commercial construction, in contrast to residential. The industry argued that there had been less speculative building, (i.e. without pre-selling to tenants) than during the 1980s boom, for instance. This view suggests an overbuild in houses during 2004-6, responding to the bubble in prices, but commercial real estate investment stayed low. Even at the peak in 2008, non-residential construction as a percent of GDP was still below the level seen for most of the 1990s. The two areas which may have seen an overbuild are retail malls, particularly adjacent to the new housing estates which sometimes, in the end, were not completed, and in the hotel segment. However, in a sense, whether or not there was too much building then does not really matter, given the collapse in demand now. Unless demand for space quickly comes back, not only to end-2007 levels but catches up the growth that was expected for 2008-9, excess capacity will remain. It would take GDP growth in the region of 6% pa for the next two years to close that gap by 2011. Almost nobody expects such a rapid recovery. Even if we saw GDP growth of 4% pa it would take nearly 4 years to get back on track. Such a strong rebound is also widely seen as unlikely, though it cannot be ruled out, but still it would still leave the real estate industry in difficulties near term.

Certainly it did overfinance?

Even if over-building was limited in this cycle, there is general agreement that loan conditions became too lax and some of the same trends seen in residential financing were evident in commercial too. For example LTV ratios rose while expected cap rates (the assumed rental yield after costs) were often inflated. One technique was to include reserves to help pay the mortgage until the expected rent increases materialised. This was usually based on the expectation that rents that had been contracted during the low point of the cycle in 2001-5 would reset higher to market rates in due course. Unfortunately the recession and falling rents have dashed those expectations, while the reserves are now running out. But there is general agreement that the most difficult segment of financing now is construction loans, a large proportion of which is financing to home builders and for condo construction.

Sizing up the losses

The commercial real estate market is valued at about $6.7 trillion with $3.5 trillion of associated debt ($2577 bn in commercial buildings and $903 bn in multifamily residential at end 2008 according to Federal Reserve Flow of Funds data). As such it is about one-third the size of the housing market, which, at the peak in 2006 was valued at $24 trillion with about $11 trillion of mortgage debt. Of the $3.5 trillion in commercial loans, about $1.8 trillion was held on the books of banks and about $900 bn represented collateral for commercial mortgage backed securities (CMBS). Assuming prices overall finish 40-50% down from their peak, the total loss in asset values will be about $2.7-3.3 trillion. This is a large number but compares with our expectation of at least a $9 trillion reduction in value in the housing stock and the $6 trillion reduction in the capitalisation of the US stock market values (at 28 July levels, i.e. 975 on the S&P). The fall in commercial real estate values, therefore, is only a small part of the overall decline in wealth the US is facing.

Losses for lenders

What about likely losses for lenders? At the end of Q1 about 7% of commercial real estate loans on banks books were considered delinquent but it is early days and the number will surely rise substantially. The IMF’s estimate of potential write-downs is $187bn for CRE loans and $160 bn for CMBS, totalling $347bn (Global Financial Stability Report April 2009). Another set of loss estimates that has received wide-spread attention is from Richard Parkus of Deutsche Bank Securities, who testified to the Joint Economic Committee of the US Congress on July 9th. He puts total losses in the region of $325-380bn but more weighted towards losses for banks than CMBS. He estimates losses on CMBS at $65-90bn, on banks’ core commercial real estate loan portfolios of $120 -150 bn and on banks’ construction loans of $140 bn. The latter segment, construction loans, is expected to see the heaviest losses, in the area of 25% partly because loss-given-default on defaulted loans could reach 80-90% in many cases. Overall, he expects the CMBS area to see the smallest losses, partly because securities usually have longer maturities than the typical 5 years for bank mortgages and also because they do not include construction lending. The loss rate for bank loans calculated by Parkus is just under 16%. Finally, another estimate of the total losses has been provided by Foresight Analytics, a California-based consulting firm, which suggested a figure of $250 bn.

The take is that it would be safe to assume a range for total losses of $250-500bn. The IMF’s estimate for total writedowns in the US is $2.7 trillion so the commercial real estate component is perhaps 10-20%. Losses on CRE should turn out to be roughly comparable with losses on unsecured consumer loans (excluding mortgages). For example the IMF estimates a loss rate of 14% on the $1.9 trillion in consumer loans outstanding which would be $266bn. But the losses in both these sectors pale beside the expected losses on residential mortgages which the IMF puts at $1421 bn, ($431 bn on loans and $990 bn on securities). For reference, the government’s stress tests of the 19 largest banks, which have about $600bn in CRE exposure, used a loss rate of about 8% in the adverse scenario, much lower than the estimates above. Still, it is worth emphasising that the losses in banks will fall mainly on the regional and local banks rather than the larger banks.

New government help will be limited

During his Congressional testimony on 21st July Fed Chairman Bernanke said it may be appropriate for the government and Congress to consider fiscal steps to support the CRE industry. Later he said ideas for fresh support for the market could include government guarantees for commercial mortgages. So far government support has been limited. The TALF programme was extended to existing CMBS but only where they are rated AAA, which is not where the problem lies. Other help has come indirectly through capital injections from the TARP fund, some of which has gone to the regional banks, which are hardest hit by the CRE collapse. In addition, banks get help from accounting rules which often allow them to keep CRE loans on the books at prices above (distressed) market values, in contrast to the way securities are often marked down heavily. It might make sense to provide guarantees to help banks refinance mortgages at higher LTVs than they would otherwise accept, to reduce the number of distress sales. However this would involve risk for the government and we suspect that using taxpayer money on any scale is unlikely.

Bottom line: CRE a significant negative but housing still the main problem

The bear market in commercial real estate prices started later than in residential but is progressing faster, with prices now down about the same amount in total. Probably the main reason for this faster pace is that lending on commercial real estate has tightened much more than on residential. For housing loans other than sub-prime and jumbo, the government has ensured that Freddie and Fannie have stepped up, but there is nothing similar for commercial mortgages and banks are very cautious about new lending. We doubt whether the government will, in the end, provide substantial fiscal support but there could possibly be some kinds of guarantees put forward at some point. In our view this will not stop prices adjusting further, adding to the stress. As a result new private construction will remain low and owners of real estate as well as lenders face further pain. The losses in the banking system, particularly for smaller and regional banks, are still not fully recognised in balance sheets so the flow of smaller financial institutions entering bankruptcy will continue for some time yet. These losses of course come on top of the losses in housing, credit cards, auto loans and derivatives already expected. This will tend to keep credit tight and therefore crimp the economic recovery. But, as we have argued, the amounts involved with CRE are only a small part of the total and significantly less than for housing. How far house prices eventually fall is still the most important issue facing the US economy.