Government Debt: A Gentleman’s Wager

We are three years into a “recovery,” but labor participation is still low, gasoline prices — albeit off their highs — remain at high levels, and economic growth is clearly substandard. In support of the case for growth, significant bad loan volumes have been written off, huge unprecedented amounts of capital have been committed to support the banking system, many government loans have already been repaid, and the government has removed any ambiguity about their resolve to support the economy. So, why can’t the economy turn the corner? I’ll bet you $1 trillion that I can tell you why.

As I thought about the grossly overindebted situation the world finds itself in, I began to wonder how debt flows through GDP. Does additional debt automatically mean additional GDP? Is it somehow quarantined on balance sheets, unable to impact the real economy? Has it been used to prop up GDP, or has it harmed GDP? Or has it been successful in manufacturing real growth? Governments and central banks have steadily increased public debt levels, lowered interest rates and encouraged ever greater debt consumption by the private sector. Has this been good or bad? Let’s find out!

GDP, of course, is the sum of consumption, investment, government spending, and net exports. Consequently, any and all debt that people take on, ultimately flows through GDP in one (or more) of these four categories. Naturally, I became intrigued with the notion that debt must flow through and contribute to GDP in some way. The first place I started is to look at the total debt in the United States over time. This includes both public- and private-sector debt. Everything. As we’ve discovered in the global debt crisis, private-sector debt can quickly shift onto public-sector balance sheets during a crisis. Moreover, when it comes to GDP growth, it’s the total economy that grows, and it’s the total debt that impacts the economy in some way. So, total debt is what matters, not just private or public debt. Consider the graph of total debt by quarter in the United States (through 2011), below.

U.S. Total Credit Outstanding ($ in Billions)

Sources: St. Louis Fed, U.S. Federal Reserve, CFA Institute.

Go ahead. Marvel at it in all its splendor: $38.3 trillion (as of 4q 2011). That’s trillion with a “T.” Done marveling? No problem, I’ll wait a moment while you try to wrap your mind around it. I know you are still a little light-headed now, but we must move on. Now let’s think about what that debt means? By itself? Insufficient. So, here’s debt relative to the size of the economy.

U.S. Total Debt as a Percent of GDP

Sources: St. Louis Fed, BEA, CFA Institute.

As you can see, over the past 40 years, total debt in the United States has gone from 130% in 1970 to 250% of GDP in 2011. Yeah, but how much did the increase in debt itself boost GDP?

Let’s start with “who.” Who borrows all this money? Borrowers come in three categories: businesses, consumers, and government. Businesses borrow money for lots of reasons. Chief among them are to build or expand infrastructure, to grow (either organically or by acquisition), or to meet operational cash flow needs. Naturally, they also borrow to replace depleted or worn out capital to simply maintain the status quo. Likewise, consumers borrow to purchase longer-term assets like homes and cars. Like businesses, consumers borrow to repair or replace worn out capital (e.g., furnaces, water heaters, etc.). Unlike many businesses, consumers also have a nasty little habit of borrowing just to consume (i.e., credit cards). Businesses and consumers together are known as the private sector. The private sector will be the basis of our analysis. Lastly, the government borrows. Governments borrow to finance fiscal deficits and — in Keynesian theory — grow the economy. So what is the net benefit of all this borrowing? How much does leverage amplify historical growth? And what is in store for the future?

In economics, all the interesting stuff happens on the margin. So, let’s look at the absolute change in total debt over time and compare that with the absolute change in GDP for the United States on a rolling five-year basis.

U.S. Trailing Five-Year Change in Total Debt and GDP ($ in Billions)

Sources: St. Louis Fed, BEA, CFA Institute.

As illustrated in the graph above, the five-year change in total debt peaked in thefourth quarter of 2008 at $12.02 trillion, while the five-year change in GDP was only $2.67 trillion for the same five-year period. However, the change in debt captures the net change over that time period, while the change in GDP captures only the increase in GDP (segment E in the following exhibit), so it is an apples-to-oranges comparison. Because the incremental debt is used, at least in part, to make investments and because it can flow through GDP in any of the years (segments A, B, C, D), not just the last year (segment E), I computed a variable called incremental GDP Flow which is the summation of all the incremental changes in GDP during each five-year period. Conceptually, it is computed as illustrated by adding segments A+B+C+D+E as seen in the graphic below.

Incremental GDP Flow

For example, if GDP increases by $500 mm in year one and by another $500 mm in year two, the incremental flow is not $1 billion, but rather $1.5 billion as the increase in GDP in year one also flows through in the following year as well. Then comparing incremental GDP flow to the absolute change in debt, we can examine how efficiently the debt flows through the economy. This “efficiency ratio” shows the proportion of total debt that flowed through the economy (not to be confused with the efficiency ratio commonly used by banks in examining their fixed costs). If the ratio is above 100%, then the debt usage was efficient as it helped the economy expand by more than a like amount of debt. If the ratio is below 100%, the debt curtailed economic growth as the debt was more than offset by inefficiencies and capital destruction. Of course, it is possible that debt accumulated in a particular period — across the entire economy — went solely into replacing worn out capital or was simply held on balance sheets thereby offering no long-run contribution to GDP growth. But wouldn’t we have been reading endless headlines about the Great Capital Replacement Cycle? Wouldn’t we have seen a dearth of investment activity? In fact, we haven’t. So, such potential criticisms seem implausible. Moreover, the fact that debt has interest costs, albeit at low rates right now, sees to it that debt doesn’t sit idle on balance sheets for very long. Actually, that’s the point of the analysis: We want to capture the net effect of adding debt over long periods, which also captures the capital replacement cycle as well.

U.S. Total Debt Efficiency

Sources: St. Louis Fed, BEA, CFA Institute.

What’s interesting about this graph is that debt efficiency started turning downward in 2003 — well before the mortgage/housing bubble peaked in 2006 — and has remained below 100% ever since. Alarmingly, it continues to worsen.

Next, let’s look at the role, if any, the Federal Reserve has played in stimulating debt and the GDP by setting interest rates.

U.S. Fed Funds vs. Change in Total Debt, Three-Year ($ in Billions)

Sources: St. Louis Fed, U.S. Treasury Department, CFA Institute.

The red shaded area at the top shows high rate environments (Fed Funds > 8%), while the green shaded area at the bottom shows low rate environments (Fed Funds < 5%). The reason to highlight these two rate environments is that one might expect growth in debt to be more sensitive to interest rate changes in more extreme rate environments. To confirm this, I then performed an event study on rates and debt growth over the past 40 years. I expected growth in debt to move in the opposite direction of interest rates in general and to be more pronounced in the extreme rate environments. Here is what I found.

Sources: St. Louis Fed, BEA, CFA Institute.

The study turned out as expected. At the extremes, movements in Fed Funds rates do have a big impact on total debt growth. But, as noted earlier, debt efficiency is declining markedly. And we are currently in an extreme (low) rate environment. Consequently, one might expect debt consumption to accelerate. However, what we observe is that government debt consumption is accelerating, while private-sector debt consumption is essentially flat. So, given today’s extraordinarily low rate environment, why isn’t private debt growing like gangbusters? Why is there now a break from the past?

To help answer this question, let’s look at the composition of total debt.

United States: Total Debt — Public and Private ($ in Billions)

Sources: St. Louis Fed, U.S. Treasury Dept, CFA Institute.

We can see that government debt has grown substantially, much like private debt, but it’s difficult to make out the nuance from this graph. What is the real story with government debt? To answer this question, I performed some statistical analysis.  This analysis clearly shows a strong correlation between private sector debt and GDP, while exhibiting essentially no correlation between government debt and gdp.  *** If you have an aversion to statistics, feel free to skip the following paragraph:

We performed a regression analysis of total debt growth on GDP growth. In this case, the regression line (the single line that best represents the data when plotted on a graph) has a good fit (based on an r-squared value of 43%). Then we performed a multiple regression analysis, in which we broke out  private-sector debt and government debt separately. In that case, the regression line has a significantly better fit. More importantly, private-sector debt growth showed a strong positive correlation with GDP, while the government debt growth showed essentially no statistical correlation to GDP growth. The r-squared value of the model increased to 50%, and the correlation of private-sector debt registered a robust 54%, compared to an absolutely meaningless correlation of just 2% for government debt.

So, now that we’ve identified the source of the problem, let’s isolate it. Consider the following chart showing the absolute change in U.S. government debt relative to the absolute change in GDP over a trailing five-years basis.

United States: Dollar Change in Federal Debt and GDP (Trailing Five-Years)

Sources: St. Louis Fed, U.S. Treasury Department, CFA Institute.

As illustrated, for the first time in at least the past 40 years, the absolute change in federal government debt has exceeded the change in GDP by a substantial margin. As of the fourth quarter of 2011, the absolute change in federal government debt was $6.5 trillion, while the corresponding change in GDP was only $1.7 trillion. However, unlike the prior analysis, we are now assuming that government debt is the only thing that helps the economy grow — getting the full benefit of any increases in debt from the private sector (to be fair, private sector debt peaked in the first quarter of 2008 and has been flattish ever since — not truly delveraging). In any event, on a trailing five-year basis, private-sector debt exhibits positive growth, so there’s no need to worry about the impact of declines. However, like the analysis with total debt, we are comparing a stock and a flow. So, using the incremental GDP flow, we not only find that the efficiency of government debt is going down but that there is a sharp inverse correlation — meaning that more government debt harms GDP. Even adjusting for the cash out the door to support the banking system through TARP and related programs (excluding the Fed), there is still $3.7 trillion figure that is unexplained.

United States: Dollar Change in Federal Government Debt vs. Debt Efficiency

Sources: St. Louis Fed, BEA, CFA Institute.

What I didn’t appreciate in the heat of the debt crisis of 2008 is the extent to which the U.S. government would increase its balance sheet. When looking at the Fed’s massive financial support, low rates, and money printing, I had feared inflation would ramp up relatively quickly. What I hadn’t appreciated then, but do now, is the extent to which the inflationary forces of money printing and Fed liquidity actions would be offset by the deflationary forces of the additional debt and inefficient government spending. The ongoing ramp up in debt, debt service, and poor allocation of capital stymies the economy and prevents equilibrium from being achieved.

Because the policies the United States is following today resemble so closely the policies followed by Japan in the aftermath of their bubble bursting in 1990, Japan offers us a window into our future, albeit with caveats. Consider first Japan’s ramp up in total debt over time.

Japan: Total Debt as a Percent of GDP

Sources: World Bank, CFA Institute.

Now, let’s look at Japan’s absolute change in total debt and GDP on a rolling five-year basis.

Japan: Five-Year Absolute Change in Total Debt and GDP (Yen in Millions)

Sources: World Bank, CFA Institute.

As you can see, the changes in total debt are vastly greater than the change in GDP, even today. Total debt was first spurred by the private sector (due to low rates by the Bank of Japan in the late 1980s) and then elevated by government debt as the private sector attempts to deleverage post 1990.

Japan: Five-Year Absolute Change in Government Debt and Private Debt (Yen in Millions)

Sources: World Bank, CFA Institute.

So, what happened to Japan’s debt efficiency as the government stepped up its debt growth?

Japan: Five-Year Absolute Change Total Debt and Percent Efficiency (Yen in Millions)

Sources: World Bank, Ministry of Finance, CFA Institute.

When isolating just Japan’s government debt, we see an even more startling correlation characterized by three pretty distinct periods. As Japan was ramping up government debt usage, efficiency decline from 1965–1979. Then, from 1979–1990, growth in government debt leveled off and so did efficiency. Then debt levels ramped up dramatically after the bubble burst in 1990 and efficiency once again plummeted.

Japan: Five-Year Absolute Change Government Debt and Efficiency (Yen in Millions)

Sources: World Bank, CFA Institute.

I will examine this phenomenon in a number of other countries and expect to find the same story intact. So, you want know where GDP is headed? Look no further than growth in private-sector credit and federal government debt. Until the strategic reasons for public and private debt change, GDP “ain’t going nowhere.” As investors, we must invest with what we expect to happen, not with what we think policymakers should do. And as for the wager, fellow Content Director Jason Voss, CFA, has kindly offered to pay the $1 trillion reward should I lose the wager. What a gentleman!