E Forecasts And Recoupling

Writing for his firm Llewellyn-Consulting.com, a former senior economist of the Organisation for Economic Cooperation and Development considers forecasts. He sounds somewhat like the late Yogi Berra about forecasts about the future:

“Many important decisions – macroeconomic, policy, corporate, military – have to be taken in respect of a future that is intrinsically unknowable. But there is no ducking the issue: decisions have to be taken. Even the decision not to take a decision is a decision. So: how best to go about taking decisions that will come into effect in, or even affect, the future? Basically there are three possibilities.

“One possibility is to forecast what that future will be. But it is a mistake to place much weight on [any forecast number] when the absolute error for OECD economies is o[n] the order of 1 percent.

“At turning points, when accurate forecasts would be the most useful, errors tend to be even bigger: the year-ahead OECD-GDP forecast made in Dec.2008 was 3 percentage points, and following the first great oil shock of 1973-74 it was 4.

“This is not to say that forecasting is pointless. A forecast assessment provides a consistent quantitative framework for thought, in which outcomes are conditional on identified judgements. Moreover, forecasts [give] the forecaster early warning when something is going on that is not understood. But there is no avoiding the fact that economic forecasts generally offer a poor guide to the future.

“Moral: base your decision on a forecast whenever you must, but look at past errors. To the extent that forecasts often do not provide an adequate basis for decision-making, an alternative is to assess the probability of different possible outcomes or decisions; and multiply that risk by an estimate of the potential consequences.

“Assessing the cost of potential consequences can be difficult, but fruitful. For example, commonly heard on trading floors before the 2008 crash was ‘The chances of this happening are only one in a million.’ Even true, however, the appropriate riposte would be ‘But what if it did happen?’ If the answer was ‘catastrophe’ then it would be wise to avoid the action.

“Moral: no decision should be taken without a risk assessment of the potential consequences. “Unfortunately, these two issues become [greatly] more difficult when the system is complex. Models of even moderate complexity, when shocked, often exhibit extreme, even chaotic, behaviour in ways that could not have been forecast and were not expected ? including by the person who built the model. “Examples are legion. Recent ones include the developments after the invasion of Iraq; and the cascade of financial and economic events following the collapse of Lehman Brothers.

“Moral: avoid allowing key variables in complex systems to go outside historical experience.”

Yesterday’s market closed barely dented by the renewed drop in the oil price, leading some experts to say this new linkage was now ended. The decoupling was ended in spades, with markets tumbling.

Moreover today Chinese stocks and the reniminbi rose but it made no difference. It snowed as people headed off to their home villages from Guangzhou (formerly Canton.)

Before turning to our stock portfolio I wanted to let you know that the website I was one of the first contributors to, www.TalkMarkets.com, now has over 500 authors. Contributors are paid in equity shares, but I also invested money to become an angel investor for the site.

It has now closed a major partnership deal with Thomson Reuters which will host all its headlines (for clicking to) at no cost to TalkMarkets. It also won the “Best Stock Site of 2015” award.

The startup is now publishing over 100 articles/day and has 17,000 registered users, including some of you. Now it has to boost its editorial team and needs more angels. If any readers want to find out more, please send me an e-mails (my phones are down) and I will put you in touch with TalkMarkets so you can become an angel investor too.

More from Britain, Israel, Italy, India, Canada, France, Brazil, The Netherlands, The Dutch Antilles, Belgium, and Finland.

Oi, Oi, Oil and Energy

*BP plc fell 2.4% yesterday. Today it careened downward some more after adjusted Q4 earnings fell 91% to $196 mn, over 70% below the Bloomberg analyst average.