A Dismal Beginning
From December 30 to the end of the first week trading week of January, the DJIA has declined by roughly 7.7% (approx. 1,370 points) and the SPX by roughly 7.6% (approx. 160 points). This was nothing compared to the mini-crash suffered by China’s stock market early this year, which continued this morning with the Shanghai Composite (SSEC) declining by another 5.33%. The SSEC has put in an interim peak at approx. 3684 on December 23 and has since then fallen by slightly less than 670 points, or about 18%. Most of the decline occurred in the first week of January.
Although the sell-off in the US stock market was comparatively mild, it still ended up as the worst first trading week of January in history. Had January started out on a positive note, the mainstream financial media would have been full of reminders of how bullish a strong showing in the first week of January was, and what a good omen it represented for the rest of the year. Instead they felt compelled to tell us why a weak start to the year should be ignored.
The contortions went as far as one analyst informing us last week that the sell-off was actually happening “in a parallel universe”. As our friend Michael Pollaro has pointed out to us, central planning worshiper Steve Liesman told CNBC viewers last week that the terrible trade numbers (both imports and exports kept declining) were actually a positive, because they would “subtract less from GDP” – as imports have declined at an even faster pace than exports. Such unvarnished nonsense can only spring from the fevered minds of Keynesians and Mercantilists.
As US imports are falling even faster than exports, the trade balance is held to be “improving” – which bolsters GDP (as we have often pointed out, GDP is a nonsensical number in many respects – this is one of them). However, a rapid decline in both imports and exports is actually not a sign of economic strength, but the exact opposite (leaving aside the fact that the balance of trade is definitely not an important yardstick of a nation’s prosperity).
Liesman apparently also argued that the weakness in assorted economic data releases should be ignored because employment is so strong. Apart from the fact that it isn’t really – it is easy to report improvements in the unemployment rate when labor force participation collapses to levels last seen 40 years ago and people hold multiple jobs that end up being double-counted – employment is a lagging indicator. If anything, this is the data point that should be “ignored”.
It was also widely recommended that one shouldn’t worry about China’s troubles, although everybody has insisted for years that China represents the “world’s economic locomotive”. Mr. Liesman also asserted that the 0.3% decline in wholesale inventories argued for “further economic strength” – evidently not considering that sales fell by an unexpectedly large 1% at the same time, widening the inventory-to-sales ratio further, which argues for the opposite.
Naturally, viewers were told that manufacturing should be ignored as well, since the services sector remains strong. However, as we have often stressed, this is a mistake, given the large share of gross economic output represented by manufacturing. The recent weakness in manufacturing data is by itself not yet a “recession guarantee”, but it is definitely not something that should simply be ignored.
Let us get back to that dismal first trading week though – how meaningful is it?
Past Examples of Early January Weakness
Normally both the end of December and early January are seasonally strong time periods in the US stock market. The market’s behavior late last year and early this year was definitely out of character in terms of the seasonal patterns observed over the past 30-40 years.
In that sense, the action in the first week of January is meaningful: strength would have been normal, and therefore information neutral. Weakness isn’t, and therefore represents a warning sign – this is presumably all the more relevant the more pronounced the weakness is, and this year a new record was established.
Thinking back, we recalled two past examples of notable weakness in early January off the cuff, one more famous than the other, but both showing patterns that could turn out to be helpful as rough guidelines for what to expect. The first example is the US stock market in 1962. There are several parallels to today’s situation from a technical perspective. Here is the chart of the DJIA from September 1961 to January 1963 (h/t to our friend BC who supplied us with the data):
Dow Jones Industrial Average, Sept. 1961 to Jan. 1963. it is worth noting that similar to 2015, the market went sideways in a consolidation near what was then record high territory during most of 1961. After pulling back sharply in early January of 1962, the market recovered a bit and rallied into a secondary peak in March of 1962. Then disaster struck, and the market fell sharply into a low in late June 1962. The March 15 peak was at 724 points, the June 26 low at 536 points – a swift decline of 26% in slightly over three months. The weakness in early January clearly was a warning sign. Today an equivalent percentage decline from the 29 December interim high in the DJIA would amount to approx. 4,600 points – click to enlarge.
The important part of the pattern we wanted to point out in the context of the 1962 chart is the somewhat weakish rebound from the January low to the March high that preceded the strongest part of the overall decline.
The somewhat more famous example we remembered was Japan’s Nikkei Average in 1990, when the bubble of the late 1980s burst. Once again we can observe a very similar pattern: a very weak showing in early January, followed by a weak rebound/ consolidation, and a strong decline thereafter:
The Nikkei Average from October 1989 to late December 1990. In this case the rebound, resp. consolidation after the early January decline was even weaker and didn’t last as long as that in the DJIA in 1962. The subsequent decline was of the same size as the 1962 decline (slightly over 25% in the Nikkei’s case), but even more fast-paced. Moreover, after the initial decline the market suffered a second down leg into an even lower low in October. It fell from a secondary high of 37,500 in mid February to 28,000 points on April 1 (losing the same percentage twice as fast as the DJIA in 1962), recovered to 33,000 points by May, and tumbled to 20,000 points between mid July and early October. At the October low, the Nikkei had lost slightly less than 19,000 points from its peak at the very last trading day of 1989, or approx. 48.5% – click to enlarge.
While we have only looked more closely at these two precedents (early January declines are actually fairly rare), they do suggest that one needs to carefully watch how the initial rebound out of a January low develops. To this one needs to keep in mind that the seasonally strong period actually continues until mid July, and is especially strong between January and May. Below is the 37-year seasonal chart of the SPX from our friend Dimitri Speck, an expert on seasonality:
A 37 year seasonal chart of the S&P 500 – the market should advance strongly between January and May.
In this sense, the weakness in the first week of January is less important than what happens from here on out. If the market doesn’t manage to rally convincingly from the January low, but instead just delivers a weak consolidation that ends at another lower high, one should probably expect to see significant weakness developing from an interim peak that could be put in sometime between February and April.
Market internals remain weak, complacency is very pronounced (see also the latest Barron’s Roundtable which had not a single Wall Street strategist making even a mildly bearish forecast), and money supply growth, while still strong by historical standards, is less than half of what it was at the peak of the Bernanke inflation.
All of this is happening against the background of growing economic weakness and a budding currency crisis in emerging markets. Overnight the South African rand (ZAR) was clobbered for 10%, falling to a new record low (it has since then recovered much of this loss). The rand is the preferred hedging vehicle of investors exposed to emerging market currency risk, as it is one of the most liquid EM FX markets.
As a result of this, EM volatility is often especially noticeable in ZAR. It also makes it a leading indicator of the entire EM currency class. Given that the currency is severely oversold by now, it may well recover in coming weeks, which could go hand in hand with a recovery in stocks from the recent sell-off. If such a recovery eventuates, its quality and character needs to be closely watched.
We have chronicled the stock market’s internal weakness and numerous warning signs such as the decline in margin leverage from a record high and the extreme complacency evident in long term sentiment measures in great detail in Q4 of last year. Now that the market has declined from a secondary lower high following the recovery after the warning shot of late August 2015, the question is whether this can provide an additional edge.
We believe it can, mainly because it is contrary to the usual seasonal pattern. To this one must keep in mind that deviations from bullish seasonal patterns are a typical beginning bear market trait. Note in this context also (h/t to Mitch Zacks for the info) that only four of the last 22 election years were bearish for the stock market (1932, 1940, 2000 and 2008). Normally it is the year after the election that turns out to be the weakest for the market.
However, as you can see, the pattern has been notably different in times we would regard as secular bear market periods. The wild ups and down in the stock market since the peak of the last secular bull market in 2000 have been accompanied by the US money supply growing by about a factor of four. This huge amount of monetary inflation since 2000 explains why the market has streaked to new nominal highs in every iteration of the cycle since 2000 (from 2008 to today, the broad US money supply TMS-2 has grown by 116%, since 2000 by 309%).
This does not mean that the secular bear market phase is over. One ingredient that has so far been sorely missing is extreme undervaluation, which the market has never reached since the 2000 peak. However, every previous secular bear market in history – regardless of whether it was accompanied by low or high interest rates – has eventually ended with market valuations at an extreme low (read: with average trailing P/E ratios in single digits). This year could therefore turn out to be yet another disappointing election year for the stock market.
Charts by: StockCharts, Acting Man, TradingEconomics, Dimitri Speck