Monday, November 9, 2015

Gold, Automation, Oil and deflation (Revised)

First off, Gold's been off its feed for some months now.  Its not clear what's going on, but in USD gold is off nearly $200 since spring -- where it had been trading around $1,200/$1,250 for nearly 3 years.  Normally, I would blame the strong dollar, but its been more or less unchanged for two years, so that's not it.  My instinct tells me that something else is a foot!  My best guess is inflation expectation are going down the tube.  Not only are primary ressources suffering (e.g. oil prices, steel etc) but the central bank's pumping action seem to achieve little.

This morning a very prominent hedge fund declared that "the floor on oil prices had been reached" within 20 minutes of this pronouncement oil price dropped nearly 8%... goes to show "nobody knows nothing".  But a massive failed auction of used heavy equipment in Australia, CEO of Maersk declaring that trade was far worse than the market assumed all point to further economic slow down. What's going on in the Oil business is a massive inventory build-up in non-conventional channels (e.g. ships).  VLCC parked off Singapore is "standard" its easy to store there, and there are half a dozen large markets within 3 days sail, but off the coast of Texas?  Everywhere there are those tankers full of oil; the trade was obvious, as the market assumed a recovery in oil price, the old adage of "buy cheap and sell dear" was too much to resist by all these hedge funds and private equity firms looking for an easy profit, not only a crowded trade but a trade that's going to wrong way!   This is driving the inflation expectations down (and there goes gold prices...)

Over the weekend two studies came out on the impact of automation; now we are not talking of auto-painters or windshield installer we are talking learning machines that can do complex tasks -- like driving a car!  The two studies one from BoA the other from McKinsey have different prospectives -- McKinsey says this will change the work environment while BoA says its the end of the world (or words to that effect).  

Two anecdotes:  When I returned to Montreal working in the trading room the F/X desk was large (although this was a small bank), it had about 7 traders actively working their clients; today there are only two traders left, the reality is that most treasurer's F/X interactions have been taken over by machines; the treasurer checks his future f/x requirements (hedging or otherwise) and the systems (from different banks) look at the projected cash flow and offer a hedging solution.  For the treasurer it allows to manage the position weekly, daily or hourly without having to justify his changes/requirements to anyone -- its a machine.  The banks lost a massive information channel to the clients, but improved margins, fewer errors and greater profits.  The trader pool shrank from 7 to 2 -- a net 5 job losses (forget about back and middle office reconciliation error rates with machine orders has dropped dramatically.  There jobs too have been eliminated.

A friend ran a business with many suppliers across the globe and a single buyer -- its the wine business.  Every day they would receive bills from the suppliers that required reconciliation, these bills were ALWAYS identical.  over a 7 year period the font changed but little else.  The business employed a team of 10 to reconcile these invoices (each product in each business and each currency) with discount information for volume.  I pointed out to this friend that for $10,000 a year the business could purchase an intelligent system that would automatically scan emails for invoices, download the invoices select the information and create worksheet and upload the data directly to accounting system -- checking twice for errors.  Total annual cost savings.... nearly $500,000.  Error rate fell to insignificant levels, reports were automatically sent to vendors -- in their own language.  Of course 9 people lost their jobs -- these were mind numbing and the staff turnover for this segment was nearly 200% per annum, error rates were nearly 5% -- which created friction with the suppliers.  However, these jobs were a net loss to the economy, even if they were boring, it was also a filter to find new long term employees -- that is now gone.   This friend saw an immediate increase in profitability, reduction in expensive errors and satisfied customers.  

The reality is that two two examples were for very different jobs; a trader earns $400,000 to a million a year, and these systems made their knowledge worthless -- it could all be coded into computers.  While the win broker's job were mind numbing they were still jobs that someone had.  The benefits to society of these trading systems was unequal, the bank saved errors and massive salaries, the client go a better system (really it was a massive improvement).  The win guy increased his profits (massive), reduced errors.  The economy lost anywhere between 17 and 30 jobs.  Potentially pure salary deflation both for high and low paying jobs


Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home