The Probable Correlation of Jobs and Office Space Demand

As I had mentioned in an earlier post, it appears that as much as 84% of the net office space demand since 2003 in the Richmond MSA may be coming from the health and education employment category. While difficult to prove without a shadow of doubt, I have outlined my analysis showing a very strong correlation of certain types of jobs and office space demand, that suggests 136 square feet of space is used by one worker in the studied categories thus indicating the influence of health and education employment.

The U. S. Bureau of Labor Statistics (BLS) publishes monthly employment data for all of the metropolitan statistical areas (MSA), of which I track nine including the Richmond, Va. MSA and that information since 1990 can be found at my web site. Additionally, the BLS tracks this employment data to a fairly detailed degree regarding type of employment, such as, manufacturing, trade, financial, leisure, health, education, government, etc. Twelve of these categories are analyzed and charted and can be seen by going here and clicking on each link.

In my opinion, the major employment categories that may produce office space demand include the following, (the links take you to their respective description at the Census Bureau's web site):

Information
Finance, Insurance & Real Estate
Professional, Management & Business Services
Education & Health Services
Government

Not all of these jobs use office space, some may be in warehouse, flex, even retail spaces, but these broad categories likely contain most of the possible office workers, whether in speculative 'for lease' space, corporate-owned, medical office buildings, or government owned properties. Since the ability to drill down to additional levels of detail in each employment category is limited by the data made available for the Richmond MSA by the BLS, the possible correlation of jobs and office space demand has to be generalized. Accepting this very large limitation, I have outlined below my analysis using historical data supplied by various sources.

At the end of 1989, two local brokerage firms Harrison & Bates and Thalhimer were the only two sources for commercial market data and each published their annual market reviews with each product type's results for that previous year. They only tracked those properties each firm believed were part of the "Speculative" market, that is, those owned by investors. They did not publish research on the owner-occupied, government, or medical office buildings and accordingly, the total market was relatively small compared to the whole universe of office space in the MSA. A table showing their year-end 1989 office statistics is shown below.

CoStar, a commercial real estate database service, tracks all office buildings in the Richmond MSA and has been doing so since 2005 on a quarterly basis. Currently, they show total office space in the Richmond MSA of 52,776,114 square feet in every type; owner-occupied, speculative, government, and medical. This is the total for the office buildings where they have the year built. Through 1989, CoStar shows a total of 35,977,781 square feet of office space as having been completed.

Therefore, all things being assumed as accurate, and using the average of the two brokerage firms totals above, we can conclude that at the end of 1989, 18,791,762 square feet of office space was non-speculative, (this being calculated as 35,977,781 - 17,186,019).

Since I know that these two brokerage firms would have included any non-speculative office buildings if they were available for lease by the end of 1989, it would be logical to assume that all of the space not in their inventory at that point was 100% occupied.

Again, accepting all of the foregoing premises, we could say that at the end of 1989, the Richmond MSA had total occupied office space of 33,140,978 square feet.

The BLS shows the following employment figures for the office related categories in January 1990 and 2009 for the Richmond MSA.

I used January's figures for two reasons; first, the BLS's figures using the new North American Industrial Classification System, (NAICS), only goes back to January 1990 for the Richmond MSA and second, this month may not be seasonally impacted by the holidays and year-end accounting, etc.

If we now divide the total occupied square footage at the end of 1989 calculated above by the total number of office related jobs in January 1990, we will get 136 square feet per worker being used on average.

33,140,978 square feet / 244,500 employees = 136 square feet per worker

If this ratio holds true, then at the end of 2008, we should have had a total of 46,430,400 square feet of occupied office space. This being calculated as:

341,400 employees x 136 square feet = 46,430,400 square feet

CoStar's data shows that at the end of 2008, Richmond had 2,250 office buildings with a total rentable area of 52,765,153 square feet, of which 4,699,367 square feet were vacant, thereby indicating that 48,065,786 square feet were occupied.

This equates to a margin of difference of 3.4% towards the predicted results over a 19 year period. Pretty close to being right on the button!

At the beginning of this post, I suggested that as much as 84% of Richmond's net office space absorption since 2003 has been as a result of the health & education employment category's job growth. With the results of the analysis above, I showed a very high correlation of office square footage needed by the five major employment categories, whether that space was speculative, owner-occupied, government or medical. This being established, we can now go back to January 2004 and compare it's employment figures to those of January 2009 and determine what percentage relationship health & education job growth had in relation to the total net new jobs created in that 5 year period. The results are shown in the table below:

Based on the net total change in office related employment of 23,600 jobs, Education & Health employment represented 84%, (19,800/23,600 = 84%) and likely resulted in the net absorption of 2,692,800 square feet of office space, (19,800 employees x 136 square feet) over 5 years.

I want to reiterate that this analysis may only apply to the Richmond area by virtue of it's employment base/office market. It may not be as accurate applied to other markets. However, I'm confident that this ratio should be applicable in evaluating Richmond's office activity over several years of analysis. Caution should be used on short term analysis, due to the typical length of office leases, delay in expansion/contraction of employees, etc.



Mike Lowry

Historical Market Charts 1989 - 2008

Since I gather huge amounts of data and statistics, much like many in the commercial real estate arena, I thought a summary of two of Richmond's longest running data collections would be useful. These two firms, Thalhimer and Harrison & Bates, have generally been collecting and publishing their data since around 1989. In years past, before the advent of the web, these firms published annual market reports and freely gave them to the various practioners of commercial real estate including their competition.

I collected these reports over the years and have summarized their findings since 1989, along with an average of their results. These are in the form of charts and can be found on my web site at www.mikelowryco.com/market-reports.html

These charts have been completed for Industrial, Flex, Downtown Office, Suburban Office and Retail properties and include vacancy rates, rentable area and net absorption on an annual basis. You will see that in many instances, their information is quite a bit different from each other. Much of this is due to which properties were being tracked by each firm prior to CoStar's entry into the Richmond market in 2005. As to the differences following CoStar, it is simply a matter as to their company's decisions as to which properties comprise their given market as both are currently using the statistics provided by CoStar.

With regard to retail properties, Grubb & Ellis|Harrison & Bates decided to drop statistical coverage of the market after 2003 excepting the vacancy rate. Accordingly, the average rentable area and net absorption, and Thalhimer's stats, track exactly in the following years.

If you review the downtown office market absorption history, you will note an ominous long term trend from approximately 150,000 square feet of net absorption annually to no net absorption. This would suggest to me that any leasing occurring in newer properties are as a result of a true lateral move of tenants from a lower class of space to a higher class. To the extent that there are firms growing and taking additional footage, there are equal amounts of space being abandoned due to downsizing, movement to the suburbs, or simply going out of business. You may think that this simply reflects current economic conditions, but, the last major commercial real estate collapse was during the early 1990's when the analysis begins so the comparisons are quite relevant.

Looking at the suburban office space charts, the change in data gathering from CoStar's involvement is quite dramatic in comparing the two firms rentable areas beginning in 2005. Again, this is due to Harrison & Bate's excluding owner occupied and government office space, whereas Thalhimer includes all. It is fascinating however, that even though their rentable areas are 24 million square feet different, their net absorption tracks nearly the same. How can it be? Owner occupied and government owned properties rarely see change in occupancy. Properties that are owned by third party investors, REITs and institutions, (typically referred to as "Speculative Properties"), see most of the action from tenants coming and going.

The Richmond area's fastest growing job category is health and education, and this has been the case for most of the past 20 years. Go to my employment charts showing this and the other major categories to see for yourself. Although CoStar tracks medical office buildings, it doesn't go far enough in providing statistics as to this employment category's impact on the Richmond area office market. Granted, not all of the health, and much less of the education jobs use office space. However, the vast majority of the finance, insurance and real estate; business and professional; information; and government employment categories do occupy office space, whether it is leased or owned. Yet, if you review these employment categories percentage growth over the past decade, most have actually been losing jobs. This being the case, the mostly positive office space absorption may be coming predominately from health and education job growth.

I'll spend some more time on this topic in a future post.

Sincerely,

Mike Lowry

Commercial Real Estate Valuation Example

Ever since learning the concept of the "Band of Investment" theory for developing a capitalization rate while attending Dr. Jack Boykin's 1977 appraisal class at VCU, I have been amazed that more commercial real estate professionals aren't aware of it.

In it's most basic form, it is simply the weighted sum of the mortgage constant and the desired equity return expressed as:

Cap Rate = (Equity Return Rate x Equity to Value Ratio) + (Mortgage Constant Rate x Loan to Value Ratio)

In a practical example, lets assume that an investor's expectation is 12% for their return on invested equity. Furthermore, his lender has assured him of their ability to provide him a 65% loan on the fictitious income producing property at 8% for 7 years with a 25 year amortization. Based on this, the mortgage constant is 9.26% rounded.

Therefore, the equation now looks like this:

Cap Rate = (12% x 35%) + (9.26% x 65%)

Cap Rate = 4.2% + 6.02%

Cap Rate = 10.22%

The pricing/valuation issues being discussed and covered in the headlines with respect to a collapse of commercial real estate values by as much as 40% currently, and the coming impact on banks which hold approximately 50% of the outstanding loans, can be easily understood if we develop a very simple example of a hypothetical property, the changes in financing over the past 2 years or so, and utilitizing the Band of Investment theory of capitalization rates.

Let's say three years ago, Investor A bought a strip center. At the time, he had banks and others competing to lend him money at 6% for 5 years at interest only. Investor A was obviously happy about this as the stock market hadn't been doing much for him and even when he was investing in stocks, the most he could borrow was 50%. The lenders competing on giving him money for his strip center acquisition were willing to finance 80% of the price. Personally, he thought that an 14% return to him on his 20% down payment would be great, but the competition to buy the center was putting him in a position to only get 8% on his initial equity.

Based on all of the foregoing, Investor A ended up buying the strip center at a cap rate of 6.4%, which was calculated:

Cap Rate = (8% x 20%) + (6% x 80%)

Cap Rate = 1.6% + 4.8%

Cap Rate = 6.4%

Now the same Investor A wants to sell his strip center to Investor B. However, in this new lending environment, the few lenders even willing to lend at all are quoting a 60% loan to price, 8.5% interest for 5 years with an amortization of 20 years. Investor B is nervous about whether all of the tenants are going to make it in this economic environment and therefore wants at least 15% on his equity.

Therefore, the new cap rate that Investor B is willing to pay at is 12.25% as calculated below:

Cap Rate = (15% x 40%) + (10.41% x 60%)

Cap Rate = 6% + 6.25%

Cap Rate = 12.25%

Not looking good for Investor A in this scenario.

A cap rate is simply the blended return an investor is willing to pay to achieve the return objectives of both his equity and the lenders loan payments. This is divided into the net income of the property to determine the price to pay.

A simple example is, $100,000 in net income is worth a $1,000,000 to someone with a cap rate of 10%.

Now lets compare the dire situation of our Investor A's strip center deal assuming that it had a $200,000 net income in 2006 and due to various reasons, that is close to the same net income he is selling to Investor B.

Here are the results:

Investor A bought at a cap rate of 6.4%, so if there was net income of $200,000, he paid $3,125,000 ($200,000 divided by .064). Of this purchase price, he borrowed $2,500,000 (80%) and put down $625,000 (20%).

Investor B wants to pay him today at a cap rate of 12.25% on the same net income of $200,000. This means Investor B wants to give Investor A $1,633,000, rounded, ($200,000 divided by .1225).

The value has dropped from $3,125,000 to $1,633,000 in three years, which happens to be around a 48% drop.

Investor A has lost all of his equity, and now he's in the hole to his lender to the tune of $867,000 if he has a recourse loan. This is more than his original investment.

Guess what? He's not going to sell....he can't.

As long as the property continues to cash flow and the lender doesn't demand payment on his loan, which they can't until 2011 for all intents and purposes, Investor A's motivation will be to squeeze as much as possible out of the rents remaining after deducting his debt service to the lender. Out of $200,000 in net income, Investor A gets $50,000 per year. This will likely mean that any maintenance, roof leaks, pot holes, etc. that Investor A is responsible for and can put off and get away with, he will, as the cash saved from not performing these tasks will flow right into his pocket.

Once this starts happening, the tenants begin to also suffer, (their customers are starting to think the center looks trashy and it appears to be in decline and are going elsewhere to shop), so some of the tenants, who are also stretched thin, start asking for relief on their rent or just stop paying at all. The issues continue to escalate as Investor A gets more desperate because he needs to get something back from his $625,000 investment and since the lender is out of town, they will not see the deterioration. Besides, his loan is likely non-recourse so he won't have to do anything but turn over the keys when he's gotten as much as he can out of the situation and not pay back the deficiency between what the property would sell at and the outstanding loan balance.

Granted, much of the story above is dramatic to show a possibility, but I believe we will see a great deal of this occurring, much as it has in previous downturns.

The biggest outstanding question in all of this remains....who is going to blink first, the owner, lender, tenant or buyer?

Home Price Index

For several years now, the S&P/Case-Shiller Home Price Index has been published for 20 metropolitan areas across the country. The data for 10 of the markets go back to 1987 and Charlotte, NC is one of those. On my web site, http://www.mikelowryco.com/ I track various trends that compare Richmond, VA to eight similarly sized MSAs including Charlotte. To see the charts I developed from the most recent Case-Shiller Index, go to this link.

Since Richmond and the remaining MSAs were not included in the aforementioned index, I was delighted to find historical data for all of my MSAs at the Federal Housing Finance Agency (FHFA). Accordingly, I have completed and posted the charts here. Additionally, I included trend lines to reflect a 3% and 4% annual appreciation from the first quarter of 1980 for each MSA, the 9 market average, and for the United States as a whole. Since I grew the index by a quarterly compounded rate, the actual increase each quarter was .742% and .985% rounded.

Based on this long term average appreciation of either 3% or 4%, it would appear that all of the reviewed markets are still quite overvalued except perhaps in Memphis, Birmingham and Louisville, (if you use the 4% annual rate). However, even if you believe 4% is a good long term appreciation rate in this FHFA Index, several markets, including Richmond are well from reaching their pricing bottoms as summarized below:

Richmond.....13% overvalued at 4% and 35% overvalued at 3%
USA.....12% overvalued at 4% and 34% overvalued at 3%
9 Market Average.....9% overvalued at 4% and 31% overvalued at 3%
Hampton Roads.....28% overvalued at 4% and 46% overvalued at 3%
Charlotte.....10% overvalued at 4% and 32% overvalued at 3%
Raleigh....5% overvalued at 4% and 28% overvalued at 3%
Jacksonville.....20% overvalued at 4% and 40% overvalued at 3%
Memphis.....23% undervalued at 4% and 7% overvalued at 3%
Nashville.....12% overvalued at 4% and 34% overvalued at 3%
Birmingham.....2% undervalued at 4% and 23% overvalued at 3%
Louisville.....0% overvalued at 4% and 24% overvalued at 3%

Now I realize this is a fairly simplistic approach towards attempting to understand what may have to happen for a complete correction in home values is over, but it may provide some insight as to how far prices need to fall before they are back near a hypothetical long term trend of between 3% and 4% annual appreciation.

Announcing Mike Lowry Company's Blog

I’m quite pleased to announce the new look and information available at my web site, http://www.mikelowryco.com/

It now contains the largest aggregation of resources, data, charts and links specifically relating to commercial real estate in central Virginia. The information about the Richmond MSA has been assembled and translated into easy to understand charts, tables and links. This readily confirmed information, from sources deemed reliable, is being presented without a bias to show how the Richmond Metropolitan area is doing.

However, through this Blog, I may provide occasional commentary as to my take on various topics regarding commercial real estate, economics, demographics, employment, housing, etc.

Thanks,

Mike Lowry