- This blog has a copy of all header posts that I publish anywhere, so that those interested in seeing what my thoughts are on the markets can find them easily.
- I will be answering questions and responding to comments, so feel free to respond to any posts and I will see your comment even if it is not on the most recent post.
- If you're interested in seeing any intraday charts I post, I do that on twitter, and my twitter handle is @shjackcharts.
- The charts in the posts are as large as I can practically make them. if you would like to look at one more closely, click on it, and the link will take you to a larger version at screencast. If you click on that again, you will get a full page version, and can use the resizing function on your browser to enlarge parts of interest further.

Tuesday, 18 February 2020

Peering Through The Fog Around Coronavirus COVID-19

This is an unusual post for me, in that it is not about the markets, but rather about an issue that might be a strong influence on the markets over the next few months, namely the current coronavirus outbreak that has so far been mainly confined to China. I'll be reviewing what appears to be the cases so far from both official and unofficial sources, and share what I'm watching to assess whether this is likely to remain a mainly Chinese issue, or might spread much wider to become a worldwide one.

Regardless of the numbers though, some aspects of life in Wuhan continue as normal. Here is a recent image of a local resident out walking their dog in a familiar scene that might currently come from any number of large cities in China:
Setting the doubtful reliability of the official numbers aside for the moment, what we do have is the daily numbers for confirmed cases over the last few weeks, so what do those tell us? Well the earliest numbers I have are from 21st Jan, at which point there were 448 officially confirmed cases and nine deaths. As of this morning 18th Feb, using the excellent live John Hopkins CSSE page, that has increased to 73,451 confirmed cases, and 1875 deaths. That's a period of 28 days in which time the confirmed case numbers have so far doubled 7.2 times, on average therefore doubling every 3.9 days or close to quadrupling every week. Obviously that's a very high rate of increase, and, if sustained, could put confirmed cases close to a million cases within two weeks. The rate of increase for deaths over that period has been marginally higher, but in effect can be considered to be the same.

What about the fatality rate? Well so far if the death rate is simply divided by the confirmed cases rate, then that comes out at about 2.5%, though an adjustment to account for the lag between a case being confirmed and death, together with the effect of the high rate of increase in confirmed cases that might put the fatality rate at a less reassuring 5% to 10%. For reference that's coming under SARS at about 10%, and the more virulent Mediterranean version MERS at about 35%. A more alarming calculation would be a division of reported deaths by reported recoveries, which comes through at about 14%, but that is likely to be an unreliable number for three reasons. The first of those reasons is that a lot of cases are mild and aren't likely to enter the reported numbers, and that also most likely applies to both the SARS and MERS numbers. The second reason is that reportedly at least, only about 14% of the confirmed cases become severe cases and not all of those will be dying. The third of those reasons is a bigger issue that applies to all of the coronavirus statistics that we have seen so far, and that is that all of the official coronavirus numbers that we have seen from China so far are likely to be very understated, in effect more propaganda than genuine statistics.

Why is this likely? Well there is no free press in China of course, and social media is restricted and monitored, but with some local VPN connections to twitter accounts and local WeChat posts there is quite a bit of leakage. The images and videos of chinese police and soldiers apparently welding shut the entrances to some infected housing developments and tower blocks are disturbing, and the WeChat employment advertisements for extra help in funeral homes and for workers to collect bodies, as well as videos showing those bodies being collected, are also disturbing, and suggest both that the numbers of deaths are considerably higher than the Chinese government currently wishes to admit, and that the situation in Wuhan particularly may have deteriorated well beyond the ability to gather decent statistics even if there was any desire to do so.

There is some evidence that extra help has also been drafted in from other provinces to help handle the extra workload at Wuhan crematoria and funeral homes, and if that's right, then the death rate must be a lot higher than advertised, as Wuhan has a population of 11 million, and the extra two thousand deaths officially reported so far wouldn't be an inordinate increase over the 10,000 to 20,000 deaths that a city of that size should see in an average month. There's also quite a lot of unofficial evidence that supply chains across China are breaking down, and that's not just an issue for Apple of course, that's also a big issue for the local production and distribution of food, which may be a serious humanitarian issue brewing down the line.

In terms of the economic impact on China this year that's all pretty worrying but the likely and ongoing uselessness of the official statistics from China should no longer be an issue for weighing whether the virus outbreak is likely to expand into a worldwide problem. Why? Well the chinese numbers may be unreliable but there are now significant outbreaks in three areas where the numbers should be of high quality, and as of this morning those areas are Singapore, with 81 cases, Japan, with 74 cases, and Hong Kong, with 61 cases. These countries all have decent health systems, and all three, with the possible exception of Hong Kong if pressure is applied on them, should report honest numbers about the progress of the outbreaks in their areas. We should be watching those carefully, and if this going to be a global problem, then that risk should first become apparent there.

What is going to happen to markets if this virus starts to develop into a global pandemic akin to the pandemic in 1918/9? Nothing good, so this is something I'll be watching with keen interest, and will be including updates as they come in on my usual market posts. I'm planning to do one of those tomorrow.

Wednesday, 5 February 2020

The January Barometer

When I looked a few weeks ago at the stat for the first five days of the year I found that the correlation with the yearly close there was essentially random, so I was wondering when I had a look at the full January Barometer stat whether I would find the same but was pleasantly surprised when I did not.

The numbers I was looking were from 1950 to the present, so that's a good statistical sample, and in that time 72.86% of years closed green, and 27.14% of years closed red, with 61.43% of Januaries closing green in that time and 38.57% of them closing red.

I've looked at the numbers for continuation from those January closes, so in the event that January closed green, then a continuation would be that year closing higher than that January close, not just closing green for the full year, and the same for bearish continuations after a red close in January. On that basis the odds of a bullish continuation into the yearly close were 83.872%, well up from the average 72.86%, and where a January closed up over 2% then that strengthened further to 87.1% odds of the year closing up from there.

In the event of a red close in January, that delivered 40.74% odds of continuation down from there into the close for the year, up almost 50% from the average but obviously still significantly less than even odds of a red close. Now January this year closed red of course, but very marginally, only by 0.2%, so I have also had a quick look to see how many red January closes were 0.5% or less, and how they performed afterwards. Unfortunately there were no other instances so I expanded that to closes less than 1% or less and there were four. That is too few to be a reliable number really but for what that's worth three out of those four closed red, for an average decline of 17.5%, helped along by one of them being January 1974. The other closed up less than 2% in 1984.

Overall the January Barometer is a solid stat that has delivered good results in the past and may well continue to do so in the future. On to the markets.

As it is likely after today's gap up and go that we are looking at a reversal candle this week, that puts the historical odds from the last ten years at about even between continuation up to resume the broken weekly upper band ride or resumption of last week's downtrend in the near future. For reference that would involve at least a near miss of the weekly upper band soon and that is currently in the 3385 area. A hit there in the next week or two could deliver a continuation of the strong uptrend over the last few months though I'm doubtful about that. I think the odds favor a resumption of last week's downtrend after this spike up has blown out, though likely we will need to see a new all time high on SPX before that, which was missed by 0.19 handles at the high today, though a new all time high was made on ES after the RTH close.

SPX weekly chart:
On the daily chart SPX closed over daily upper band today and we may well see the retracement we didn't see today tomorrow instead. There is some support for that on the shorter term charts.

SPX daily chart:
On the SPX 15min chart my channel resistance trendline was broken by the gap up this morning, and that has expanded into an even nicer rising wedge resistance trendline. If that survives tonight, unlike yesterday's resistance trendline, then then the obvious next move would be a test of the rising support trendline from Friday's low, which closed today in the 3305 area and is rising at about four handles per hour in regular trading hours. If that wedge support breaks then this move might be over, though I don't think that any SPX high here that failed to make a new all time high would last long. I can't recall any instances offhand from the last decade at least of a high on SPX lasting more than a couple of weeks where ES had made a new all time high and SPX had not.

SPX 15min chart:
On the futures charts I have decent quality possible 60min sell signals brewing on ES, DAX and ESTX50, though obviously they would need some kind of a retracement to fix, and that negative RSI divergence might be lost soon if the equity futures continue falling upwards the way they have the last two days.

ES Mar 60min chart:
On NQ and RTY those 60min sell signals have already fixed and both of the last two retracements including last week's move started with a 60min sell signal on NQ alone, so I'm watching that with particular interest.

NQ Mar 60min:
My lean is towards more downside and that's only in part because I'm not seeing any evidence that the coronavirus situation isn't likely to get a lot grimmer over coming weeks, if only because the virus incubation period almost guarantees that the confirmed case numbers are likely to get at minimum an order of magnitude larger.

That being said, there is never any certainty in market action and this has been a very persistent uptrend, perhaps because (not officially) QE4 has been pouring record amounts of liquidity into markets since October. That's still ongoing and China has added another $250bn of liquidity and stimulus to that this week.  That's a lot of money and it that's certainly been having a strong impact since yesterday's pre-market announcement from China delivered the strong gap up yesterday morning.

On Friday I'm planning the third post this week crunching the coronavirus numbers and considering the virus news. I'll be doing a market update then too, so by that stage we should have a better idea whether any retracement is possible on equities this week other than the impressive retracement we finally saw on TSLA today.