Semantic Sophistry

The mainstream financial media has been trying for years to redefine words like “correction” and “bear market”.  Unsatisfied with the traditional qualitative usage in the field, it first added an arbitrary 10% threshold in order for a decline to qualify as a correction, and a 20% threshold for a bear market.  As you might expect, trying to shoehorn a continuous analog real world into a binary digital way of thinking has led to more confusion than clarity.

We cynics suspect there may be an agenda behind this.  The above may seem innocent enough, but more recently the pedantic framework has evolved so that a market is down 20% before it is said to “have entered” a bear market.  Could it be that this helps avoid usage of a term that might scare some of the hoi polloi out of their stocks while they’re still high?

Then this summer some began insisting that the market was “in correction” until it regained the lost 10%.  This led to headlines touting the length of the correction that started back in January.  This then leads naturally to comparisons with prior corrections of a similar length, along with a resulting bullish conclusion.

Today MarketWatch ran a story exquisitely highlighting the folly of going down this road in the first place.  Part of it obviously struggles with the question of whether the current bull market in US stocks will soon be the longest.  Notice how the following quote at once endorses and dismisses rigorous application of the 20% bear market criterion:

A bull market is seen as ending when stocks fall 20% or more from a cycle high. The market’s pullback from July 16 to Oct. 11, 1990 was only 19.9% on a closing basis, so sticklers would argue that the move should be classified only as the “deepest of corrections, not a bear market,”

https://www.marketwatch.com/story/the-longest-bull-market-in-history-not-so-fast-2018-08-15?mod=hp_minor_pos21

Much ado about nothing.  Whether we call a decline of 19.9% a bear market or not changes the size of the decline not one iota.  The only possible use for such might be to support a view as to what the market may hold in store for us going forward.  And if it turns on a mere 0.1% difference in a particular historical period, is there any reason to trust it?

Another story navel gazes into the deep question of whether a correction is over until a new high is reached.

Dow DJIA -0.54% stands now about 2.8% short of emerging from correction territory after hitting its 2018 closing low of 23,533.20 on March 23. It needs to close at 25,886.42 or above to achieve that. Some market-technician purists believe that an asset must put in a new high to officially emerge from correction phase. Other technicians say that a 10% gain from an asset’s low is sufficient to exit correction territory, a characterization that MarketWatch adheres to.

https://www.marketwatch.com/story/dows-200-point-tumble-puts-it-at-risk-of-the-longest-stint-in-correction-territory-in-nearly-60-years-2018-08-15

It seems that the opinion of market technicians is now objective fact, reported as if the infinite variety of patterns displayed by markets can be sifted and categorized into distinct phyla whose behavior can be accurately pigeonholed as qualitatively different in kind by virtue of small quantitative differences in magnitude.  We have this all figured out folks!  If the Dow would just close above our magic number it’s all up from here.  Insight through lexicography.

Needless to say, here at Financology we won’t indulge in that kind of sophistry.  For our purposes here, we will stick with tradition and consider any decline extended in time a “bear market”, and any reversal of a trend having gone too far in one direction a “correction”.  A “bull market” is any rise extended in time.  A “crash” is any large decline taking place in a short period of time.  We do not “enter” a bear market when it is down 20% … a bear market begins at the top and ends at the bottom.

We won’t try to turn quantities into word games.  If the exact size of a market move is relevant, it’s easy enough to put a number on it.