"While land sales comps have been relatively scarce in recent quarters, for ease of illustration we assumed that land prices have fallen by 50% from the peak in 2007," Twist says. "We also assume hard costs are down approximately 15%, reflecting lower material and labor costs." While construction costs are down, cap rates have risen by nearly 200 basis points from sub 6%. With a cap rate of 6, developers could have built to a yield of 6.9% in 2007, generating a 15% profit margin. But in order to generate the same profit with today's cap rates, developers would need to build to a yield of more than9%, according to Twist. "Interestingly, the required rent today is very similar to what it was nearly three years ago, given the falling costs and increasing cap rates ($65 cost x 6.9% equals $4.49 in 2007, relative to $50 cost x 9.2% equals $4.60 required rent today)."

However, market rents are down substantially. And most rational developers would not build until rents, costs and yields made economic sense--or were moving decisively in that direction. According to Twist, it is important to note that today's replacement costs will likely begin to grow again over the coming years, thereby increasing even further the rents required to build.

Production, trade and inventories are the principal drivers of demand for industrial real estate, together explaining more than 80% of the variation in historical demand. As such, the respective forecasts for these leading indicators can be used to estimate the future magnitude and timing of demand for industrial real estate. "When we analyze these variables in conjunction with the pipeline of new construction and estimates of new supply, we can forecast the availability rate for the US and individual markets and submarkets," says Twist. "With 2009 experiencing the first contraction in global GDP and the steepest decline in global trade, demand for industrial space hit a record low. However, the current trend and consensus outlook for trade and production, when input into the AMB trade and production model, imply the US and the world should realize more than 100 million square feet and 500 million square feet of net absorption in 2010 and even more in 2011.

If the current consensus estimates and forecasts for production and trade are realized, AMB's model also implies the overall US market would likely see rental rate growth as early as 2011, with equilibrium reached by 2012. So as the US market moves toward 10% availability over the next two years, rents can be expected to reach levels necessary to support new construction. According to AMB data, rents will grow by more than 30% over the next three to four years. The company has developed replacement cost, forecast, equilibrium and rent growth models for markets and submarkets globally. "The rent growth trajectory varies by market and submarket, affected by the spread of current rent to replacement-cost-justified rent, the level of availability relative to equilibrium and the forecast of demand and supply," says Twist. "As mentioned, the coastal markets have likely hit an inflection point in terms of demand and there is virtually no new supply being constructed.

The leading indicators of demand for industrial real estate, such as production and trade, are clearly rebounding, Twist says. And the outlook for these indicators for 2010 and beyond is even stronger. "We expect that these improving economic conditions will fuel recovery in demand, starting in the infill markets. The consensus forecast for global trade and production suggests that more than 500 million square feet of demand could be realized globally in the next few years, driving the availability rate to equilibrium levels in 2012." All of this should drive market rent increases to levels necessary to support new construction requiring several double-digit spikes in the coming years.