The Phillips Curve just might hold the answer to the worst-case scenario of low growth and high inflation, says OMAM's Stewart Cowley.

Economics really is a dismal failure. It started out as a neat mental set of images that helped people understand the world. You were never really expected to find evidence for these mental pictures let alone start to run countries on the basis of it. In many respects, things were fine for economics while the facts were, by and large, hidden. The collection of data on a wholesale basis was for a very long time blissfully difficult to do. Testing economic theory against what actually happened remained a minority activity.

A case in point is the so-called Phillips Curve, which, plainly put, says that, as unemployment increases, so inflation decreases. The idea is logical - a reduction in demand because of increasing unemployment causes producers to reduce prices to bring the system back into equilibrium, and the fillip to the economy helps kick-start the cycle all over again. It's a nice sensible picture. The trouble is, it doesn't work.

Take, for instance, the Phillips Curve for the US. It looks more like someone has taken a shotgun to a barn door. You COULD get a computer programme to draw a straight line or polynomial curve through it, but it would be a fanciful argument that led you to positing any kind of deterministic relationship. Data collection has debunked a rather nice idea.

But this is too harsh. Interestingly, the relationship varies from nation to nation and on what length of time of data you include. For instance, Japan shows a startling coincidence between theory and practice over the long term. The question arises as to why it is not universally applicable at all times and in all circumstances and to all nations.

The key seems to be, 'the larger your deficit with the rest of the world, the more chaotic your Phillips Curve'.  In other words, the more you have an external deficit (as much as to say a dependency on other people's money), the more your currency will flip around. In turn, this means your domestic inflation rate becomes more chaotic and less related to your correcting unemployment rate as the Phillips Curve would dictate.

Japan has a very strong external balance and a strong Phillips relationship. The UK and US have large external deficits and terrible Phillips Curves. Europe once had a large positive balance with the rest of the world and a Phillips Curve that was respectable. Lately, the region has slumped into deficit, explaining why, even with unemployment of 9% across the European region, inflation remains at 3% when it "should" be less than 1%.

The consequence of deficit reduction, the new mantra of the West, will be low growth and high unemployment. Believers in the Phillips Curve are expecting inflation to fall rapidly and even deflation to set in. But since we are now systematically in deficit and at the sway of the currency markets, this is not a foregone conclusion. In fact, the opposite may happen as our currencies decline. At the very least, inflation will stay much higher than policymakers are expecting or markets are currently anticipating given the level of unemployment.

For instance, over the next 10 years, US inflation is (according to the markets) expected to average just 2%. In the UK, the figure is 2.6%.  These numbers are ridiculously low if we have lost control of our own domestic inflation rate.  The combined effects of chronic external dependencies (both in terms of deficit funding and trade), the fragility of our currencies and the ability of emerging nations to buy the basic stuff of life from underneath us because they have money in the form of reserves could send inflation materially higher or keep it at the levels we are experiencing today. In anything other than the magical realism of some financial commentators, investors should be compensated for this with higher nominal yields and higher implied inflation rates from the markets. Neither of these things is happening at this time. The lack of compensation is just not something that can go on forever.

From our point of view, there are things we can do about this. Buying inflation-linked bonds in the US and UK are the obvious outlets, but, should 'real' yields rise, just buying inflation-linked bonds won't be enough to create a positive return. Any rise in yields will cause index-linked bond prices to decline because even an index-linked bond possesses a duration - the rise and fall of prices as yields fall and rise.

Instead, we want access to the inflation aspect without the duration effect. To do this, we have been buying inflation-linked bonds and selling government bond futures against them. This way, as inflation expectations rise, the combined pair of a long and short makes money in an environment that is normally hostile to bonds. Conversely, if inflation expectations fall, the pair will experience a loss.

So far, this strategy appears to be working - in the US at least, where inflation expectations are rising. We suspect we will be doing more of this trade in the future, but there's plenty time yet.

Stewart Cowley is head of fixed income at Old Mutual Asset Managers.