Bonds In A Bubble?
We keep reading all these articles about how bonds are in a bubble and that any day now the bubble will burst. Naturally when you see something going straight up, the contrarian in each one of us wants to say "Hey, that's not right". Many felt that way in the late 1990s when the NASDAQ was in its bubble with price earnings ratios over 100 and even as high as 200. So out came the short sellers, convinced that they would make a bundle when the bubble burst.
The trouble is that bubble markets can go higher and last longer then the early short sellers can stay solvent. Once they are forced to throw in the towel they contribute to the bubble with massive short covering.
And so we have today's bond market, with rates falling to levels never seen before in our lifetime. So are bonds in a bubble?
Unsurprisingly, many economists deny that bubbles even exist. Long-term sharp rises in markets are merely a part of the boom and bust cycles that have existed for eons. So there is no clear definition of a bubble. There is just no real theory out there. But huge spikes in markets have been observed for centuries, going back to the days of the famous mania in Dutch tulip bulbs in 1637. It seems that greed, the greater fool theory, following the herd and of course excessive monetary liquidity remain the prime causes of bubbles.
The recent sharp rise in bonds is a function of a sinking deflationary economy and a rush into the safety (or at least the relative safety) of government debt instruments. In a word - the fear factor being the opposite of greed. You could hold cash, which is the ultimate IOU, paying no interest, or stick your funds into government debt instruments such as Treasury Bills or Treasury Notes and Bonds - all now paying next to no interest. With the Federal Reserve lowering the official bank rate to zero to 0.25 per cent, terms of two years and less all pay under one per cent. From overnight to one year it is darn near zero. In other words government debt instruments are now much the same as holding cash.
Canadians are suffering the same fate. Government of Canada bonds carry low interest rate yields. Most banks pay very little to hold either deposits or on their bank accounts. Amazingly, though, it seems that the more debt the government issues, the lower the interest rate it must pay. But that won't last.
What is striking about this rush into bonds is that it is just government bonds. Corporate, municipal, state (provincial in Canada) and many other sovereign bonds have not followed US Treasury bonds down. Indeed, even as government bonds rallied, many corporate bonds in particular fell in price as yield spreads widened. The weaker the credit, the more the spread widened. So if there is a bubble it is specifically in government bonds and more so in the US then here in Canada as the rally in Canadian bonds has lagged their southern cousin.
While it may be argued whether bonds are in a bubble, they are probably reaching the top (or bottom in yields, as bond prices move inversely to bond yields) of a very long cycle. The last cycle of high yields topped in 1981. So this current up cycle in bond prices has run for 27 years. Prior to that bond prices were in a long-term decline dating from 1941, when long US Treasury bond yields last bottomed at around 2.0 per cent. This of course came during the Great Depression. So rates rose for 40 years and have now been falling for 27 years.
This long-term chart of US Government Bonds is quite striking (courtesy www.sharelynx.com). Since a clear peak in rates around 1815 (War of 1812-14), rates fell for about 15 years to 1830. Rates then rose from roughly 1830 to another peak around 1862 (Civil War). Then came a very long decline as rates did not bottom until about 1901 (39 years). Rates then rose for another 19 years until 1920 (WW1). The next down cycle lasted roughly 20 years until 1941, bottoming as WW2 got underway.
Quite tellingly, the long rise in rates from 1941 until 1981 culminated as a result of another war (Vietnam) that helped ratchet up the inflation rate. Long-term rates are truly a function of inflation itself and there is often nothing better than wars to boost inflation (and commodity demand) and interest rates, with the resultant spike in the demand for financing to help pay for those wars (to date the Iraq war has cost $587 billion and it is still rising. Afghanistan is another cost). For years now we have had the War on Terror in Iraq and Afghanistan and elsewhere. Add to that the potential trillion-dollar deficits and inflation could once again rear its ugly head.
Trying to discern the pattern here is difficult as periods of falling rates lasted 15, 39 and 20 years, and the current cycle is in its 27th year. Periods of rising rates lasted 32, 19 and 20 years. One could argue that the long up cycle from 1830 to 1862 was really two cycles, with the first peaking about 1848 (Mexican/American War) and the second peak around 1862, but what all of this tells us is that trying to determine long-term cycles in interest rates is fraught with danger. We may have longer to go in this current cycle but with rates falling not far from their long-term lows in 1941, the fall may not have much longer or further to go.
But it might be worth studying the current up cycle that dates from the rate highs in 1981. That's when interest rates peaked at 20 per cent. We labelled that low in price as a (in red on our chart). Secondly we note the huge ABC pattern that took it out of that low. We labelled that up wave as a Big Wave A up.
Since this is a log chart, note also that the move from the 1981 lows to that top in April 1985 is more spectacular then what we are witnessing today. Bond prices moved $41 for a gain of 91 per cent. Even the move from B to C was roughly $37.50, for a gain of 77 per cent. By contrast the most recent move from the lows in June 2007 has been just over $37; only a 35 per cent move as prices are at a much higher level.
Since then we have been moving inexorably upward in price (down in yield) in a fairly well defined channel. Since 1985 we have reached the top of the channel four times: 1993, 1998, 2003 and the recent one in 2008. We overthrew the top of the channel in 1998 and we have done the same this time. The time between each top, counting again from the 1985 high, was roughly seven years, five years, five years and five years.
We have attempted to label this long rise from the 1985 top. We counted the pattern down from 1985 that fell in an ABC fashion, followed by another up and then a down as our B wave of a larger degree bottoming in 1991. Since then we appear to be following some sort of possible triple double abcxabcxabc pattern. Whether the final ABC has played itself out from the lows in 2007 or we have a b wave followed by one final c wave up to our final high to create our final C wave of a larger degree is difficult to say.
Ray Merriman (MMA Cycles Report - www.mmacycles.com) has noted what he believes is a fairly reliable six-year cycle in bonds. We note these cycles seen in 1981, 1987, 1994, 2000 and 2007. In the last case there was an equivalent low in 2006 but for now we will use 2007 as the final low of the cycle. We have labelled this cycle as - a, b, c, d and e - all in red, following our long-term trend line up.
Each of these six-year cycles can break into either two three-year cycles or three two-year cycles. The first three phases of the cycle - a to b, b to c and c to d - all broke down into two three-year cycles. The most recent six-year cycle broke down into three two-year cycles. In his latest 2008 forecast Merriman has premised that US Treasuries are due for a cyclical top in 2008, maybe extending into 2009, and that once that top is in we could see a liquidity crisis much as we saw in the bond market collapse of 1987-88 and again 1999-2000.
Certainly we are seeing signs of a possible top in Treasury Bonds as we write, but as noted above this final phase of the topping pattern could yet subdivide into a correction followed by one more attempt to reach the top of the channel. Since we topped in December 2008 at 142^21, US Treasury 30-year bonds have fallen $10. Yields have jumped from lows of 2.53 per cent to 3.04 per cent. Ten-year yields are up from 2.08 per cent to 2.43 per cent.
If we are indeed topping then the next two-year cycle could have occurred as early as July 2008.The range would have been from February to November. That may be possible as we did experience a double bottom low in June and October 2008 falling right in the time frame. That would only occur if we counted our last six-year cycle from the low in 2006. If the cycle is counted from the June 2007 low then the 2 year cycle low is due in July 2009, plus or minus five months as noted by Merriman. And that assumes that this six-year cycle again unfolds in three two-year cycles. Based on that our sense is it is possible that we have reached a crest in bond prices but more likely is the cresting of the bond wave up has a number of months to go before it makes its final top. This then would make the next two year cycle pushed out to 2010 again centering on July 2010.
With interest rates possibly topping we need to examine what will cause interest rates to rise. We note them below.
The massive monetary growth seen since the financial crisis got underway. M1 is up 11.5 per cent year-over-year, M2 by 7.6 per cent, while it is estimated that the now unreported M3 is up over nine per cent in the past year. The most recent thirteen-week rate of growth has been even more astounding: M1 + 36.9 per cent, M2 + 13.8 per cent. All inflation is ultimately monetary inflation.
Deficits as big and as eye-popping as you have ever seen. President Elect Obama has talked almost casually of annual trillion-dollar deficits for "years to come". This is an astounding 6.9 per cent of GDP (based on $14.4 trillion for the third quarter of 2008). If the deficits go even higher, as expected, the ratio to GDP will rise further as GDP is contracting. According to The Economist the US budget deficit for 2008 is only 3.2 per cent of GDP. This would more than double it. Looking at the budget deficits of various countries, this would put the USA in the same league as Egypt and Pakistan. Pakistan is an economic basket case with its hand out to the IMF. Egypt probably is not far behind. Both are dependent on aid from the United States.
Who will finance these massive deficits? If lenders come forward it will be at the expense of other borrowers, causing massive crowding out. Corporate, state, municipal and other sovereign borrowers will find it increasingly difficult to raise money due to the huge demands of the US. And many lenders, particularly foreign lenders, are already choking on US debt. China has over $1 trillion of it. With domestic concerns of their own, and their $600 billion stimulus package to finance, and with their economy contracting, China will be forced to spend more at home rather than buying US$ denominated debt. China's voracious appetite for US bonds has helped keep US interest rates low for years. It was a quid pro quo, as the US bought China's widgets and China bought the US Bonds. Now it is about to come home and haunt both of them.
Some estimates from the Office of Management and Budget show that over the next 30 years the budget deficit could grow to 20 per cent of GDP. That we believe won't happen because the US would probably go bankrupt first. The official Federal debt is currently listed as $5.8 trillion (Source: Federal Reserve statistical release Z1 Flow of Funds) at the end of the third quarter of 2008, or 40 per cent of GDP. But total public debt is closer to $10.7 trillion or 74 per cent of GDP. The difference is because the Flow of Funds only reports debt held by the public and not the intergovernmental holdings. These levels are going to rise sharply with additional spending, bailouts and the deficits. And all of this is assuming that revenues stay afloat. In fact they will now fall because of tax cuts and reduced flows due to higher unemployment and lower income. It is projected based again on figures from the Treasury and the Office of Budget Management that total Federal Debt could rise to 250 per cent of GDP by 2040. Astounding.
The Federal Reserve will be forced to become the buyer of last resort for US debt. This means they will be printing money, which again is highly inflationary in the long run. Some have talked about it as being potentially hyperinflationary.
The deficit numbers do not take into account the huge shortfalls in unfunded commitments such as Social Security, Medicare and pension and health programs for military and civil service. These unfunded commitments total today about $52.7 trillion (GAO Financial Condition and Fiscal Future Briefing January 2008 www.geo.gov).
If war concerns increase, and we have no reason to suspect that they won't, then the deficit will rise further to finance any increased defence expenditures. The US military expenditures currently around $515 billion represents roughly 42% of the entire world's military expenditure and is larger than the next ten countries combined. The United Nations budget represents roughly 2 per cent of the world's military budget.
The US dollar has enjoyed a good run over the past few months, driven by the need to repatriate funds from foreign carry trades due to forced selling from hedge funds and others. This process is nearing the end and the potential for the US dollar to resume its downward spiral will soon return. We may not be quite there yet but we are getting close. This is another reason for interest rates to rise: to protect the value of the dollar.
Bonds in a bubble? Sure, fear has generated the flight to safety, particularly at the short end of the yield curve where rates have fallen virtually to zero. But for long-term bonds, despite the nice run up in prices (fall in yields) over the past few months the party is getting near the end. All signs point to higher interest rates further out.
This brings us to our last chart: the gold/bonds ratio. Since the bottom in gold back in 2001 the upward march in prices has favoured gold over bonds. Since the ratio peaked in July 2008, the past several months have favoured bonds over gold. But the past few weeks has seen a rebound in the ratio once again in favour of gold.
While it is clearly possible that the long-term uptrend line (currently near 5.000) could be tested in the weeks ahead, if that were to occur then the long-term uptrend in favour of gold would resume. Failure to take out the lows seen in 2006 would, however, be quite bullish. Gold has been the top-performing asset class over the past decade and we have no reason to suspect it is about to give up that mantle. Indeed, with the risks to bonds outlined above, owning gold becomes even more imperative.
Note: Charts created using Omega TradeStation. Chart data supplied by Dial Data.