$50 Billion? -- The Volatility of Deficits and Markets

The recent revelation of a sharp upward increase in the forecast federal budget deficit for the current fiscal year is getting a lot of new coverage, and a lot of disbelief over how this could happen. By 'happen', most commentators mean not the deficit's size but the quality of the government's budget forecasting, and the possibility of dishonesty in their reporting of the numbers.

While I have my own thoughts on the degree of political interference with the budget numbers going back to before the fall election, let's assume for the moment that the government is not acting deceptively in any way. There is ample room for poor predictability during stable economic periods and even more so in the present crisis. We can criticize the quality of the economists and others in Finance, yet their job is not at all an easy one. Let's start with the following extract from the summary of the budget numbers, that immediately predates the most most recent revision:


You should be able to see where the previous $34B headline deficit figure is in the 2009-2010 column (-$33.7B). It is simply the sum of the revenue and expense items above it in the column.

First, notice the relative stability of expenses in comparison to revenues. Program expenses are not static, but the rate of increase is pretty steady and would appear to be in line with growth trends in the broader economy. This is unsurprising since the government has a lot of control over its expenses. This is even true with the sudden increase in Employment Insurance (EI) payments - if it gets too onerous they could, via Parliament, change the rules to bring it down to a more desirable amount. All programs are like that. Whether they choose to manage those figures comes down to policy and politics. My point is that the government has control over these expenses.

The debt charges are modest and change by an even smaller amount when compared to total expenses. This is expected since this line item is the interest rate on the total outstanding debt. We can therefore ignore this line item for our purposes.

The fireworks is all in that very first line: revenues. To understand this line we need to recognize that the wealth in this country is almost all in the private sector. If you consider the capacity for wealth creation then it is entirely in the private sector. The government does not create wealth; instead they take a portion of the private wealth and distribute it into a variety of public institutions and programs. The best the government can do in regards to wealth creation in our system is to provide a legal environment that is favourable to private sector wealth creation.

With the understanding that the wealth that is the source of all the government's revenues must come from the private sector we can begin to see that the government is only partially in control of its revenues. Let's look at a few examples, all of which (unsurprisingly) relate to some form of taxation.

We know that a lot of government revenues come from income taxes, both business and personal, and that personal income is an expense for businesses (in the form of salaries). Consider a hypothetical company with $100M in expenses and $110M in revenues. It's profit is $10M or 10%. It pays taxes on that $10M of, say, $1M in federal tax (since I'm focusing on the feds I will not look at provincial taxes). If the company is in a knowledge industry like high-tech, the bulk of company expenses are in salaries. Let's assume $80M for this compay. The personal federal taxes paid by employees might then be about $20M (income, GST, etc.).

Federal revenues from this company is something like $21M/year, or about 5% of its revenues. This is obviously a very simplified analysis, though I believe it is adequate for my analysis. If it is a public company in a moderately stable industry, the company's stock price will reflect its P/E (price-to-earnings ratio) in some fashion. The earnings are $10M in the present case, or divide by the float to get the P/E on a share basis.

Now let's have the company hit the skids as many are doing nowadays. Business evaporates, which is reflected in reduced revenues. For example, $70M rather than $110M. However their expenses are unchanged unless they reduce salaries (their primary expense) by dismissing employees that are no longer contributing to the lowered business volume. Let's assume they choose to dispense with profit for the near term so they dismiss ~30% of their staff to bring expenses equal to revenues. Their profit is zero, so no business income taxes are paid. $30M in salaries are no longer being paid, so federal taxes from the former employees declines from $7.5M to $1M. From this we see that federal revenues have declined from $21M to $13.5M. That's a lot, so it's good that most companies are not in so dire a situation.

In addition to the lost revenues, the government also must pay EI to those displaced employees. This does not replace all the lost salaries, but it does increase government expenses, some of which are recovered in taxation (though at a lower rate since the personal income level is lower). Let's say that program expenditures increase $10M and revenues increase $1M.

What started as net government income of $21M in this example is now $4.5M, a decline of nearly 75%. In fairness, this is only a short term effect since economic forces will adjust to the new circumstances, such as the rise of new industries, recovery of the old, and re-employment of the displaced. But since we're looking at this one budget year, the drawn scenario is instructive.

Now, let's get back to this distressed company with no earnings. Its stock price tanks (we've all seen this happen) along with P/E. The company is now less able to raise funds by debt or equity, reducing its ability to retool its business or make other adjustments to its operations. Its stock, much of which is in the hands of pension funds or cash investors, suffers accordingly. If the company goes bankrupt, as is likely with GM, the shareholder wealth is reduced to zero, forever.

For cash investors, there are capital losses instead of capital gains, which further reduces government revenues. Desperate shareholders pull money out of their RRSPs, at reduced valuations, to pay the bills. This act increases government revenues since the withdrawals are taxable, but the amount is lower than if the withdrawals were at higher a share value. Everyone reduces spending, unleashing a cycle of reduced GST taxation, further loss of business taxation, and then more employee displacement. Add it all up and you get -- a $50B deficit.

Our economy is a classic non-linear, chaotic system in the mathematical sense. As in weather forecasting no model of such a system can do a good job when looking forward more than a short distance. When a major exogenous event occurs, predictability becomes a joke. Flaherty is not responsible for that.

This is already a long post and I hope it was worth the time to read. I can't truly quantify or justify the size of the deficit or its frequent revisions, but when I look at the underlying upheavals on the street I can sympathize to a degree with the government. The politics of the matter can only serve to obscure what is largely a problem originating in the private sector - the dog - where the government finances are the tail of the dog: the dog wags the tail, and not the other way around.

One bright point is that the reversal when it comes could be just as sharp. The federal deficit for 2010-2011 could be far better than what's in the table above, even with the upward revision in this year's deficit.

[Addendum: I forgot one important point, the one that ties back to the title of this post. The volatility of the deficit mirrors the volatility of the markets for much the same causes, as demonstrated in the scenario I presented. In particular, small declines in revenue cause large declines in profitability, which then eventually cascade into a sharp decline in the tax base.]