Futile Forecasting Frustration 

Having run a workshop last Saturday, called 'Men Managing Money Maturely', I am on an alliterative roll.  It is somehow addictive.

What has prompted today's blog was the simple and unassuming exercise of cycling to work.  As part of my master-plan to keep the ravages of time at bay, I cycle everywhere I can.  One of the biggest dangers in the process is pedestrians, especially those walking their dogs.  Today, my serene progress through the park was impeded by the apparently random actions of walkers - but this is far from unusual.  One needs permanent 360 degree vision to safely anticipate the behaviours of pedestrians, and I have often wondered about having one of those Google Streetview cameras welded to my cycle helmet.

Today was no exception, and as I cycled with great care past the rose-garden, I was already attempting to anticipate what was going to happen at the T-junction at the end, where my arrival was due to coincide with a diminutive lady and her dog, who were somehow taking up the full breadth of a path that is at least eight feet wide.  Which way would she turn?  There were only two options, but as she steadily approached the moment of choice, there was no indication at all of which it might be.  I slowed right down so I was matching her pace.  If she gave any indication of a left or right bias, I could take the opposite side of the pavement.  But no, right until the very last moment, it was as if her trajectory would take her straight across the T-junction, and straight into the bushes.

At the very last second, the lady swerved to the left.  The dog went right.  I veered off into a particularly thorny shrub which rather unfairly received some invective from me, of which I am now slightly ashamed.

Now, here's the thing:  how simple was that scenario?  One slow-moving geriatric.  Only two possible choices.  The whole model slowed down so the forecasting can be continually reassessed.  Surely to goodness, it would be possible to make a reliable prediction in such circumstances?  But no.  From an observational perspective, any attempt to forecast a left or right bias would have been entirely random.

Which brings me to the folly of market timing.  Jason Butler writes briefly on the subject in his useful book, 'Wealth Management: How to plan, invest and protect your financial assets'.  Other, more technical, books deal with this in more depth, but actually you almost don't need to go to all that effort.  The moment you begin to bullet-point the range of factors which can influence the behaviours of a market, or a given asset-class, or a fund management group, or a particular fund manager, or a specific investment fund, then you begin to realise that the probability of crafting an algorithm to allow for all those variables is vanishingly small.  It may be possible.  It may be that our universe is merely one in a billion parallel universes where such a thing is possible.  And pigs might fly.

Of course, most IFAs don't have in their possession such an algorithm, should it even exist.  I imagine such a thing being hidden away in a steel box, in the vaults of the Vatican, under permanent armed guard, lest IS capture it, along with Middle-Eastern oilfields.  It would be like that closing scene from 'Raiders of the Lost Ark', with the box vibrating with latent power, in some dark corner.   The evidence is that nobody has such a tool, otherwise active fund managers would do a much better job.  So, if the Vatican doesn't have it, and if the vast majority of active fund managers don't have it, what infallible tool might an IFA seek to rely on, if he (or she) is to pull off an accurate market-timing feat more than once?  Once is just luck, but if one is to set out one's advisory stall to encompass some kind of promise of added value, via this arcane wizardry, one had better depend on something more substantial than gut feeling.
 

What does your investment advice model look like?

ValidPath Members have access to a guidance page to help you work through all of the essential components of your model.
Kevin Moss, 20/11/2015