The credit crisis that began in 2007 served as the trigger for one of the most serious recessions since the Great Depression. We all know that numerous companies were severely impacted by the recession, especially those whose businesses are especially cyclical (e.g., automobile, steel, etc.). What is less well known is that many state and local governments (municipalities) have also been seriously impacted by the recession. Declining economic activity tends to simultaneously reduce tax revenues and increase expenditures as more citizens are forced to rely on the social safety net.

In the early 1990s, some well-regarded financial engineers proposed that cyclically impacted corporations and municipalities could hedge their "macroeconomic risks" by engaging in macroeconomic derivatives (such as swaps or options). A macroeconomic swap would be a swap in which one leg pays a fixed rate of interest and the other leg pays a floating rate of interest tied to some "macroeconomic index" (such as the growth rate of GDP or an inflation rate such as the CPI). Macroeconomic derivatives are often embedded in other instruments. For example, the U.S. Treasury department's TIPS (Treasury Inflation Protected Securities) contain embedded inflation swaps. In the early 2000s, Goldman Sachs and Deutsche Bank teamed up to offer macroeconomic derivatives (which they called economic derivatives) to their clients. These were structured in much the same way the financial engineers had suggested ten years earlier.

Now suppose that you run a risk management consultancy firm and you have been hired by a mid-sized municipality to help them avoid future economically-induced fiscal crises. Specifically, the municipality wants to determine to what extent their cash flows are impacted by changes in regional inflation-adjusted GDP growth rates and then to hedge that risk. To help them, you have obtained quarterly inflation-adjusted regional GDP data from Bureau of Economic Analysis (BEA). BEA reports that for the past 40 quarters real GDP for the region has grown at about 2 percent on an annualized basis. You have also obtained, from your client (the municipality), inflation-adjusted revenue and expenditure data for the past 40 calendar quarters (not annualized). You need to determine by how much a change in annualized GDP quarterly growth rates will impact the municipality's quarterly net cash flows. Net cash flow is defined as the difference between the municipality's gross revenue and gross expenditures. This data is in the spreadsheet.

You should:

1.Now suppose that the cyclical behavior of the municipality's net cash flow exposes its bondholders to credit risk. Presently, the municipality has $1 billion of debt outstanding on which it pays an average annual interest rate of 6% this reflects a premium for the credit risk. Ratings agencies have told the municipality that if they could remove the cyclically-induced credit risk, they could reduce their borrowing costs by 40 basis points (i.e., 0.40%) a year. How much would they save annually if they can remove the cyclicality of their net cash flows?

4. 

Cyclically Sensitive Municipalities (and Corporations)

At the end of August 2010, the city of Harrisburg, announced that it expected to default on its debt by missing a $3.9 million interest payment. Harrisburg is the capital of the State of Pennsylvania. Actual default was averted, at least temporarily, when the State of Pennsylvania announced a week or so later that it would accelerate some grants and other payments to the city to allow Harrisburg to meet its upcoming debt service. The bonds are insured by a municipal bond insurance company, so investors were not at significant risk, but a default would drive up Harrisburg's, and other municipalities', future funding costs. At the time of this writing, the cities of Detroit, Michigan, and San Diego, California are in similar straights.

The problem for these three municipalities, and almost all others as well, is that any economic slowdown will cause a decline in their tax revenues while increased community needs will make greater demands on the expenditure side. Almost all municipalities are cyclically sensitive. But the degree of cyclicality varies considerably from locale to locale. A city that is home to many cyclically sensitive businesses will itself be particularly cyclically sensitive. A city that is dominated by colleges and universities will be less impacted, at least for a time, by economic stress.

Given this cyclical sensitivity, one might be inclined to think the solution is simple: Municipalities should build reserves during expansionary periods in order to have a rainy day fund to draw on during contractionary periods. But history proves that few municipalities can do this. Politicians are always under pressure to show the voter that they are conscious of the voters concerns especially with respect to taxes. Surpluses are hard to justify as they lead to pressure to either (1) increase spending, (2) cut taxes, or (3) some combination of the two.

Macroeconomic derivatives could easily represent a powerful, albeit partial, solutions to this problem. Using historic data, a city like Harrisburg should be able to determine how changes in national or regional GDP growth impact its cash flows. Alternatively, it might do the analysis using the national or a regional growth rate in non-farm payrolls. The city could then enter into a GDP or non-farm payroll swap. We will use the GDP swap to illustrate the process.

Let's suppose that a city's risk management committee (RMC) knows from experience that the municipality's budget is balanced when the real economy grows at 2.7% -- which is the official estimate of the long-term real GDP growth rate. The RMC (or outside expert hired for that purpose) has further determined that a one percent change in GDP translates into $20 million of net cash flow for the city. That is, for every one percent increase in GDP the city will have an annual extra $20 million of net cash flow. (Net cash flow is the difference between revenues received and the sum of budgeted and

unbudgeted expenditures.) Conversely, for every one percent decrease in GDP the city will suffer an annual $20 million decrease in net cash flow. When the city is in a cash surplus state, it experiences pressure from special interest groups to spend more and from taxpayers to cut rates. When the city is in a cash deficit state, pressure mounts to raise taxes (always politically unpopular), spend more on social safety net programs, and borrow in the capital markets. The goal is to keep the city's budget balanced in all economic climates.

So the city enters into a ten-year GDP-swap with a macroeconomic swap dealer that uses the same estimate of 2.7% long-term real economic growth as measured by GDP. This growth estimate is used to price the swap. Suppose that they structure the swap such that the city pays the swap dealer the actual annual growth rate in GDP on notionals of $2 billion and the swap dealer pays the city an annual fixed rate of 2.7 percent on the same $2 billion of notionals. Note two things: First, the notional principal is not real money and no one gives this to the other it only exists for purposes of calculating the later periodic payments. Second, in practice, there would be a spread, measured in basis points, on the GDP leg to compensate the dealer for its role in the swap, but we will ignore that in this example. The structure of the swap together with the municipality's cash flows from taxes and expenditures is depicted in The beauty of the swap solution is that the city is now insulated (i.e., hedged) against a large portion of its macroeconomic (a form of systemic) risk stemming from the cyclicality of the city's cash flows. Importantly, the city would not want the swap to have a very long tenor because a city's internal dynamics change over time and the degree of cyclicality may be impacted by that change. To see that the swap works, suppose that the economy's growth rate increases to 3.7 percent. Then, under the terms of the swap, the municipality would pay the dealer $74 million (i.e., 3.7% x $2 billion) and the swap dealer would pay the city $54 million (i.e., 2.7% x $2 billion). In a swap, only the net is exchanged with the higher paying party paying the lower paying party the difference. So, in this case, the city pays the swap dealer $20 million which is precisely the size of its surplus for the year.

On the other hand, suppose that the following year the economy sinks into a recession and GDP growth becomes negative, say -1,3%. Then, the swap dealer will pay the city $26 million on the GDP leg. (This is because the payment on the GDP leg is negative, so it goes in the opposite direction.) And the swap dealer also pays the city $54 million on the fixed leg. Thus, the city receives an infusion of $80 million from the swap dealer thereby offsetting its cash flow shortfall caused by the decline in GDP (i.e., the recession).

One of the beautiful things about this solution to a municipality's cyclical macroeconomic risk is that it would likely remove much of the political pressure to increase spending in good economic times and decrease pressure to raise taxes and cut services during poor economic times. Thus, it fosters an environment of stability which facilitates long-term municipal planning.

Macroeconomic swaps (and other macroeconomic derivatives) are not as easily constructed as we might have led you to believe. In the swap market, dealers act as intermediaries between end users. To function well in this capacity, there needs to be a two way market. Of course, there are far more corporations and municipalities that are cyclical than counter-cyclical, so one might argue that the market cannot work. But, there are ways around this problem as has been demonstrated by the Case/Shiller real estate indexes. These indexes became the basis of futures contracts and swap dealers can hedge in the futures. It only requires that there be sufficient speculative interest on the other side of the futures contracts. There are, of course, other problems with swaps of this nature such as how to handle revisions in the GDP number which can happen several calendar quarters out. But those sorts of issues are resolvable by employing appropriate lags.

Another solution, which really involves an embedded swap, very much the same way that the inflation indexed bonds involve an embedded swap, is to structure municipal debt offerings with a floating coupon such that is tied inversely to the growth rate of GDP. For example suppose that, under current market conditions, our aforementioned city could issue a 10-year note paying a fixed rate of 5%. Instead, suppose they issue a note paying 2.3% plus the growth rate of real GDP. If the long-term growth rate of GDP turns out to be 2.7%, then on average, the city has paid an annual coupon of 5% over the ten

years. But in years when GDP is above its long-term estimated growth rate, they pay more and in years when GDP is below its long-term estimated growth rate, they pay less. This sort of debt financing would have negated the need for the city of Harrisburg to announce that it was going to miss an interest payment. Indeed, this sort of debt financing would likely lead to a long run increase in the debt rating of the city. Rating agencies would see the city's finances as more economically stable. This, in turn could lead to an even lower fixed rate component. For example, the coupon might be 2.1% plus the GDP growth rate.

In concluding this chapter, we would hope that we have demonstrated that there is lots more room for financial innovation and financial engineering. If done right, it can help address systemic problems that have plagued municipalities and corporations alike since the birth of the modern state.

Quarter 1 2 3 4 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 annualized GDP growth rate % 2.20366 1.42981 4.03151 3.17109 2.20170 1.84874 5.57684 5.51188 4.91021 1.25852 1.26613 5.31170 3.07300 4.39021 3.17530 2.07842 3.38717 3.09552 3.06665 3.03434 3.12437 2.29554 2.09879 3.97218 1.52284 0.24247 2.17267 (0.63995) 2.58036 1.27183 (1.19737) 0.48518 (1.02822) 1.32837 3.08778 2.82230 4.27398 3.53091 3.09967 3.43398 in millions in millions Quarterly Revenue Quarterly Expenditures 111.66456 116.02655 112.26292 116.85823 123.56972 109.52809 118.46415 114.84887 109.53796 116.23801 111.68625 115.20062 126.08466 107.77336 128.64243 108.69670 121.56798 111.11335 102.95193 116.93421 105.88752 112.88753 124.82363 115.39704 118.81607 110.03874 121.58414 109.57530 120.53679 111.55644 114.32824 113.77150 113.92076 112.88821 121.04717 111.99817 121.36847 110.14525 119.37140 111.50880 112.13879 113.58584 107.83114 116.00611 115.86219 112.48733 123.55634 107.97472 107.51314 118.49560 104.44569 120.15441 111.02172 113.37364 86.62401 118.98728 113.12635 113.29117 108.08772 117.29220 98.56697 122.26152 99.02059 121.81462 96.00006 122. 13690 104.90961 114.77858 124.17482 109.82724 123.99908 112.25505 127.77929 107.17163 120.12803 109.16230 104.81654 120.01055 115.37592 116.45015 Exhibit 13.5 A Macroeconomic Hedge Revenue Sources Sales Taxes Income Taxes Real Estate Taxes Cash Inflows (cyclically sensitive) %AGDP x notionals Municipality NCF = $20mx(%AGDP) Macroeconomic Swap Dealer Writes GDP Swap 2.7% x notionals Cash Outflows (cyclically sensitive) Notionals = $2 billion Expenditure Obligations Schools Emergency Services Social Welfare Programs Quarter 1 2 3 4 5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 annualized GDP growth rate % 2.20366 1.42981 4.03151 3.17109 2.20170 1.84874 5.57684 5.51188 4.91021 1.25852 1.26613 5.31170 3.07300 4.39021 3.17530 2.07842 3.38717 3.09552 3.06665 3.03434 3.12437 2.29554 2.09879 3.97218 1.52284 0.24247 2.17267 (0.63995) 2.58036 1.27183 (1.19737) 0.48518 (1.02822) 1.32837 3.08778 2.82230 4.27398 3.53091 3.09967 3.43398 in millions in millions Quarterly Revenue Quarterly Expenditures 111.66456 116.02655 112.26292 116.85823 123.56972 109.52809 118.46415 114.84887 109.53796 116.23801 111.68625 115.20062 126.08466 107.77336 128.64243 108.69670 121.56798 111.11335 102.95193 116.93421 105.88752 112.88753 124.82363 115.39704 118.81607 110.03874 121.58414 109.57530 120.53679 111.55644 114.32824 113.77150 113.92076 112.88821 121.04717 111.99817 121.36847 110.14525 119.37140 111.50880 112.13879 113.58584 107.83114 116.00611 115.86219 112.48733 123.55634 107.97472 107.51314 118.49560 104.44569 120.15441 111.02172 113.37364 86.62401 118.98728 113.12635 113.29117 108.08772 117.29220 98.56697 122.26152 99.02059 121.81462 96.00006 122. 13690 104.90961 114.77858 124.17482 109.82724 123.99908 112.25505 127.77929 107.17163 120.12803 109.16230 104.81654 120.01055 115.37592 116.45015 Exhibit 13.5 A Macroeconomic Hedge Revenue Sources Sales Taxes Income Taxes Real Estate Taxes Cash Inflows (cyclically sensitive) %AGDP x notionals Municipality NCF = $20mx(%AGDP) Macroeconomic Swap Dealer Writes GDP Swap 2.7% x notionals Cash Outflows (cyclically sensitive) Notionals = $2 billion Expenditure Obligations Schools Emergency Services Social Welfare Programs