It's Still The Economy, Stupid

Thursday, January 29, 2004  

Tax Revenues and the Housing Bubble

MaxSpeak has a pretty good post analyzing the origin of our budget deficit. Most telling is the graph of revenues versus outlays, as a percentage of GDP. Please remember that the 2004 outlays (which appear flat vis-a-vis 2003) are due to a) the rapid growth of GDP in the next 12 months, likely IMO to be overstated b) expected outlays based on the current Bush budget, which if the past is any indication, will be wildly understated.

While Max makes a macro analysis, I'd like to take a micro look using a hypothetical taxpayer and working upward.

Tax revenue peaked in 2000, coincidentally with the stock market bubble, which immediately causes many economists to assume "capital gains taxes", or the other coincident event: Bush tax cuts. Just from looking at the graph, the total rise in revenue from 1992-2000 was around 3% of GDP, revenue jumped from 17.5% of output in 1992 to 20.5% of output in 2000. If all that was capital gains related, it wouldn't nearly cover the 4.5% drop from 2000-present. The same is true of Bush's tax cuts. The deficit increased from a $150 billion surplus in 2000 to a $350 billion deficit in 2003, and the deficit is expected to be around $450 billion in 2004. The difference is around $600 billion or about 5.5% of GDP. Since the deficit is computed from both spending and revenue, that 5.5% change needs to explain both the 4.5% drop in revenue and the 2% rise in spending. While the tax cut/spending theory gets us closer to that 6.5% combined difference, it doesn't cover it in entirety (all number expressed as % of GDP).

While I feel both of these played a role in the current budget problems, another major influence is certainly the housing bubble. As housing values rose 50% from 1998-2003 (according to the Federal Reserve's Z-1 tables), mortgage debt was up more than 70%. Mortgage interest on that debt is deductable when you itemize your taxes, and (as I've found out to my delight), the yearly interest on a mortgage on even a median priced house in many metropolitan areas easily tops the standard deduction, even the modified standard deduction under the Bush tax cut (to "eliminate the marriage penalty", the deduction for a married couple was raised to double the single level).

For fun, you can play with the IRS withholding calculator to see how $1000 in extra mortgage interest works in your favor. The itemized deduction is a constant rate (your tax rate) on the dollar against withheld taxes. So, for a married couple making $125,000 a year, the 25% tax rate would mean $250 less revenue for Uncle Sam for every $1000 in mortgage interest deducted. A single person making $50K a year is in the 15% tax bracket and another $1000 in mortgage interest means $150 less for the gummint.

Let's make a very rough and conservative estimate of how much money this could be costing the government. We'll guess home ownership to 40 million U.S. households. Let's take a median mortgage of $140,000 in 1998 and 70% higher in 2003 ($238,000). Multiply 40 million versus $238,000 to check: $9.5 trillion. That's pretty close to the $9.2 trillion in mortgage debt in the third quarter, plus a bit for the gain in the fourth quarter and any mortgage debt that is not included in this line item but in some sort of "other" category.

To be extremely charitable, we'll assume an average 15% tax rate. Historically, there was a roughly 8% mortgage rate in 1998 and 6% rate in 2003 for the average household. This works out to $11,200 interest paid in 1998 (.08 x $140,000) and $14,280 per household in 2003 (.06 x $238,000). It would follow that at a 15% tax rate the average household pays $600 less in taxes due to higher interest payments and 40 million households means a total of $18 billion less revenue for the government. Since this is certainly a conservative estimate, we'll double it to put in a range $18-36 billion per year in reduced revenue due to higher mortgage interest deductions.

This estimate pales in comparison to both Bush tax cuts, which are estimated to have reduced tax revenue by $150-200 billion, the higher unemployment rate (probably adding $100 billion more in lost revenue) and spending increases ($150 billion). The insidious part of the mortgage income tax reduction, though, is that it works only after the standard deduction has been eclipsed, and once it is worthwhile to itemize, it caps income tax withholding in future years. So if someone were to project revenue for future years based on a recovering economy, the income tax projections would be underprojected by $18-36 billion plus the additional mortgage interest that would incur from a larger mortgage debt base.

To expand this analysis, note all the factors involved in calculating the impact of the estimate. If interest rates were to rise, all other things equal, a 1% increase in mortgage interest rates costs the Federal Government about $400 per household (.01 x $238,000 x .15). A 10% increase in mortgage debt costs the government about $350 per household. Lower interest rates have actually saved the government quite a bit of money the past three years by reducing the amount of mortgage interest being deducted. With mortgage interest rates at forty year lows, however, I doubt we'll see much improvement in the future.

Referring back to MaxSpeak, the projection through 2014 is for government revenue to climb back to 20% of GDP in ten years. This projection assumes quite robust economic growth rates that would surely mean higher interest rates and higher home prices than the current situation.

Based on an extremely conservative estimate of current mortgage debt growth rates (around 5% per year versus the actual 10% growth per year), a return to the 30-year average of mortgage rates (+.2% per year to 8.2% in 2014), homeownership increasing 2% per annum, the effective tax rate rising by 0.5% per year (to 20.5% in 2014, along with revenue) and our lower bound current estimate ($18 billion), the cost of higher mortgage interest deductions will be (thank you Excel) $30 billion in 2004, $44.5 billion in 2005, and finally $273 billion a year in 2014.

As if we didn't need this in addition to the recession-free assumptions of 4.5% nominal GDP growth for the next decade, an unemployment rate of 5.2% in 2014, inflation averaging 1.9% average for the next decade, etc., etc that the CBO uses in estimating tax revenue.

I hope you can deduce from all this that the assumption for federal revenue growth is hopelessly unrealistic, even under "conservative" CBO assumptions. The continuation of the housing bubble only changes the qualifier from "hopelessly" to "laughably".

But as with most of the Bush Administration policies, particularly regarding Iraq and the economy, "laughably unrealistic assumptions" are par for the course. This administration is the most fiscally irresponsible this country has ever had (and probably ever will), and with several bubbles in the balance and $33 trillion in total debt, this is certainly not the time to be unrealistic with so much at stake. The flat truth is that Bush and the GOP have crippled the federal budget for DECADES in just three years. Add to this that the crippling private pension funds, federal pension protection, state unemployment insurance funds, social security, medicare, health insurance in general, the entire social safety net is in tatters, facing enormous future obligations that have no hope of being funded under realistic projections. At this point, replacing this administration is only going to be a first step in fixing a legacy of economic troubles.

posted by Teddy | 9:26 AM |

Wednesday, January 28, 2004  

Dow Drops 142 points on Fed statement

Just when I stop paying attention blogging and stock prices, the market pulls a fast one. The FOMC changed its statement slightly, removing the part where it stated the Fed would keep interest rates low "for a considerable period". At the time the statement was issued, the Dow was up around 20 points, so I think we have some causality there, as from that point forward it was straight down and the Dow finished off 141.55 or 1.33%. The S&P 500 was down 1.36% and the Nasdaq fell 1.83%.

The only worry for this economy is higher interest rates, particularly mortgage rates. When the Fed raised rates in 1989 and 1990, instead of the "soft landing" they got a recession and an S&L crisis. When rates were raised again in 1995, we had a near recession. When rates were raised a third time in 2000, we had another recession en lieu of a "soft landing". Thus ended the Fed's attempts to raise interest rates even a little (the 2000 hikes were milder in absolute terms than either earlier period).

An economy that runs on debt and credit creation simply cannot tolerate significantly higher interest rates. While rates are falling, an economy saves less and less and consumes more and more, holding less cash relative to outstanding debt levels. But if interest rates are falling, debtors reduce their interest payments, which compensates for lower cash generation, and everything seems hunky dory. But when rates go up again, it is difficult to meet payments with the gradually restricted cash flow.

Also as a consequence, large lenders have a very simple equation determining profits: interest earned minus interest paid plus capital gain minus default. As rates are falling, capital gains are boosted, particularly if the entity owns bonds as lending reserves. It boosts the difference between interest earned and paid if the entity can refinance faster than their customers. It also reduces the default ratios as lower rates allow more credit to be accessed by customers, so assets grow faster than losses, even if losses are growing at double-digit rates.

When the 10-year treasury-bond rate jumped by 150 basis points in a month during the summer of 2003, it nearly gave debt markets a heart attack. Fortunately, either the selling subsided or a large buyer, (such as the Japanese Central Bank or the Fed itself), stepped in to buy Treasuries and the yield on the bonds topped out around 4.6%. Thirty-year mortgage rates headed up to 6.3%, but once again fell with Treasury yields. But since the jump in rates was not enough to make refinancing completely useless, homeowners simply switched to adjustable-rate mortgages and continued to access the equity built up over reent years. This would not be the case had the 10-year Treasury jumped another percentage point or two, or especially if the yield curve were to flatten slightly, which would drive up ARM rates relative to fixed-rates.

The Fed's change in wording sent shudders through debt markets today. Ten-year treasury rates jumped 11 basis points to 4.2%, up from a bottom of 3.97% set just two weeks ago. There is the smell of another blowout in swaps ahead, as bond traders usually go home earlier than stock traders do and really won't react until tomorrow morning. It will be interesting to see what happens to bond rates now that Alan and Co., ever slightly, have removed the "forever" from their "rates will stay low forever" FOMC statement. I'm sure traders had thought the statement was permanent as long as the economy wasn't generating hundreds of thousands of jobs. This has to be a shock.

Eventually there has to be a threshold where interest rates are so high that refinancing is out of the question, and the desire to buy a second, third or fourth house is no longer profitable. Now is not that time. Homeowners have such a huge return, thanks to capital gains, tax credits and low interest rates, that the only thing restricting mortgage credit and homebuying is the lenders themselves. The question is, what would drive mortgage interest rates up to that level, say 7 or 8%, and what are the risks?

In one scenario, the decline in the dollar could finally give us some CPI inflation (say 5-6% a year). Interest rates would then rise to give lenders a positive real return. I don't think this is the most likely scenario. First, the CPI is horribly biased against inflation. Second, as has been discussed earlier by Kash at Angry Bear, the U.S. is partially shielded from a weaker dollar by foreign exporters willingness to reduce profits to keep dollar prices steady. Thirdly, higher inflation would mean that fixed-rate mortgage payments would deflate (hopefully) relative to income. U.S. capital prices would also deflate, and any producer with extra money could snap up U.S. capital and avoid the increasing price of foreign-produced exports (similar to the Japanese reaction to the big decline in the dollar in 1985-87).

In the second scenario, China (or some other country), allows their currency to revalue and no longer needs to buy dollars to prevent the revaluation. This would cause treasury bond rates to rise and other rates would move in sympathy. This scenario is more likely than the first and could take place very gradually, depending on how many billions in bond purchases are cancelled. I can't really envision any country shocking the market, as this would cripple economically the chief importer of their products (as well as call attention to themselves for U.S. reprisals). Such a policy could also be cancelled or reversed fairly quickly.

In the third, and most likely scenario, the swaps market blows out like it did last summer, repeatedly, and rates keep ratcheting upward to higher levels. This seems to be the environment the Fed is fostering. I think expectations now are that sometime, the Fed will raise rates again, and every speculator in Treasuries feels they will get out before everyone else. They also rely on the Fed to telegraph their intentions, and make moves in miniscule amounts. Yet today's action shows that any movement whatsoever, even in language, is enough to precipitate a stampede out of long-dated Treasuries.

We'll see just how big the stampede is in the next couple days, but that it was precipitated by such a minor change in wording has to worry the Fed. They really don't have much room to raise rates, period Once they do, expectations will surely be for more than just one hike, and all the central banks in the world won't be able to stop the speculative deluge. While the JCB has billions, my last check of open-interest on treasury bond futures was around $1 trillion. God knows how much of the $170 trillion of derivatives function by treasury bond holdings. If mortgage rates were go get back to 7-8%, it's hard to conceive how housing demand could maintain current levels, or that mortgage lenders wouldn't start hemorrhaging money since most of their customers have locked in between 5-6%.

posted by Teddy | 2:42 PM |