"Gentlemen prefer bonds?"

As someone who is currently scraping the bottom of the fourth quartile over the past three months, the words of the song "The Only Way is Up" come dancing across your brain as you negotiate the toy train set that tries to pass itself off as the London Underground each morning. But since the dust seems to have settled on a period of time that has been as deeply disturbing as it has been unprofitable, you have to start to wonder about the process we have been through in the past six weeks.

The first thing to recognise is that the majority of rout in the bond markets has been caused because marginal buyer of bonds - the western pension fund industry and the Far East - have, by and large, disappeared. Either because equities have risen (closing pension fund deficits) or just the last duped investors had been drawn into market, by mid-January there was clearly nobody left to buy as the mathematics of the situation did not work for anyone anymore. No wonder the economists took over. After all, bond yields have been held lower than you would have expected, if you like economics textbooks because of excess demand more than anything; by our estimates about 50 -100bp lower.

In that case, It's not surprising that comparisons with other periods of rising yields have become rife in the press - everybody is always looking for the new bond bear market for some reason. For this reason alone, say the words "Nineteen ninety four" to any bond manager and you will see them trip into a dream-like state of pain and anguish as they recall one of the worst episodes for bonds in recent memory. During that year, the US Federal Reserve caught everybody on the hop and suddenly increased interest rates. The unwinding of highly leveraged positions and, admittedly some assumptions that interest rates would stay at just 3% forever, saw bond yields rise by a full 2% during the course of a single year. It was very uncomfortable for everybody. It's not unreasonable to expect people who are desperate to find something to write about, to make comparisons with that period in an effort to make sense of the situation.

For those people who like that kind of thing, if you plot the path of yields in 1994 against what has happened since the beginning of this year, then it is obvious that there is a kind of similarity in the trajectory (see first graph). The starting points have had to be re-based back to one to create the comparison, but, if you followed it through and expected a repeat of 1994, you would expect US long-term bond yields to peak at around 5.85% in the fourth quarter of this year.

Of course, as with all of these simplistic comparisons, the real question is what is different this time around? The first thing we know is that the western world is just that much more sensitive to relatively small interest rate movements because so much money has been borrowed privately, publicly and for the purposes of speculation - it is doubtful that a 2% increase in global borrowing costs could be tolerated without serious consequences. For instance, US bond yields, the housing market, employment and thereby the rest of the world, are related in a way that hardly have to be rehearsed here. But it may be instructive to understand, for those who think they can ignore their bond portfolios for the rest of the decade, why there are self-limiting processes going on here which will lead to a decline in yields in the not too dim and distant future.

As we have pointed out in these pages in the past, the US Federal Reserve cares not one jot about the niceties of university taught economics - all it really cares about is jobs. There is very good evidence to show that the Fed only really reacts when jobs are either being created or destroyed too fast. The problem for the Bernanke Fed is that, since the millennium, US job creation has been far too dependent on the housing market for comfort. In the past six years some 25% of jobs, outside farming, have been generated in the construction-related industries. This is an uncomfortable number because the construction industry has only a passing relationship to employee loyalty, but more than a kissing relationship with borrowing costs.

Employment changes in construction-related industries are incredibly sensitive to interest rates and bond yields because of their linkage to mortgages and how that can drain money from consumers' pockets (see second chart). Given half the chance, the construction industry will lay off people at a breathtaking rate as has been seen several times in the past. In this chart, we have inverted the annual change in construction employment to show that as interest rates increase construction unemployment increases. Clearly, there are time lags involved, but equally the sector cannot tolerate limitless interest rate increases before the policy bites but by the time the policy has some effect, it is too late. In other words, get interest rate policy wrong and the whole process goes into reverse very quickly.

Given that the US economy is so sensitive to interest rates, you have to conclude that what has gone up is going to come down (in yield terms). The current bout of tightening clearly isn't having any meaningful effect on the employment situation yet but it may just be the case that it is because the transmission mechanism into long-term interest rates broke down a couple of years ago creating Alan Greenspan's "conundrum". With the agents that cause this now on the side lines (the indiscriminate buyers of bonds) we may be in a position to resume normal service in which case yields should rise a bit further and Fed policy should start to bite in a meaningful way.

We know Fed policy is currently having an effect and, therefore, there should be considerable value in bond markets. The following downswing of yields should reverse some of what has happened so far this year. So, there is little doubt that, when the time comes, gentlemen will prefer bonds, but we may have to be patient to pick a new entry point and keep watching the US housing market for clues.