- WE'RE JUST RANDOM SPECKS OF DUST IN A TORNADO TO THE MARKETS .......
- CHARTISTS MUST PUT ALL BIAS ASIDE AND LET THE CHARTS DO THE TALKING OR WE'LL SEE ONLY WHAT WE WANT TO SEE
- 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, 24 March 2020

A Short History Of Market Crashes

My last few posts have been a coronavirus COVID-19 series, so I'm putting in the links here so as to refer back to them easily for now. These are the seven posts so far.

18th February - Peering Through The Fog Around Coronavirus COVID-19
24th February - Some Genuine Coronavirus Numbers Coming Through
28th February - Falling Down The Steps
9th March - A Tale Of Two Cities
12th March - Sudden Death
16th March - Pinball Markets
20th March - A Short History of Superflu Pandemics

Announcements to start with today, as apparently some people have been having trouble with the link to register for the special free public webinar that we have scheduled at theartofchart.net at 4PM EDT Sunday 29th March to look at COVID-19 and the Financial Markets, and in that webinar we will be considering COVID-19's impact on the economy, systemic risk and impact on dealers, the Fed's actions and credit risk, and reviewing potential trading opportunities. If you'd like to attend you can sign up for that here. If you are having any trouble finding a link to any webinar that we are doing you can also always check my twitter or my partner at theartofchart.net Stan Nabozny's twitter to find the link.

On to the markets where I've been promising a post looking at the obvious target areas for this market decline to end. I'm doubtful that a low is close, though I am thinking that SPX might find some kind of floor to bump around for a few months while the COVID-19 pandemic plays out and we get a clearer picture of the economic impact that is likely to have. I'm going to be looking at a couple of previous well known market crashes to illustrate why I think that the usual fibonacci retracement targets are still valid for this move as well.

The first thing to say is that not all of this decline is about COVID-19. At least some of this is about the very high level of the market before this crisis, and the degree to which that high level was attributable to the low interest and high QE environment that the Fed have been delivering since 2008/9 with the aim of that being, in significant part, to inflate stock market and other asset valuations. To some extent the fast decline we have seen is just the air coming out of the market as it was pricked by an event that was large enough to push stock markets out of the upward path that the Fed has guiding them up for years. That said, there is a real crisis here, but at least some of this has just been a much delayed reversion to the mean:
I'm going to look at a couple of famous past market crashes first as there is a widespread belief that these crashes are uncontrolled events that can go anywhere, whereas I would argue that they are generally just retracements that happened alarmingly quickly, while nonetheless heading to predictable targets that were obvious buying opportunities to the trained eye. In these times the smart money tends to do very well, while the it is the less well informed investor that tends to end up as a chalk outline on the pavement:
The first famous crash I'll look at is the 1987 crash on SPX. This has something strongly in common with the current decline in that it happened very quickly. There had also been a strong bull run into the decline, up over 200% in 5 years. What I would note though is that, despite the shocking stories about the panic and wild decline, the decline was an almost perfect 50% retracement of that bull run, after which SPX bottomed out and started the next move up. My view is that the 1987 was just a retracement that happened alarmingly fast.

1987 Crash - SPX Weekly 1980-8 chart:
The second crash I'll look at today is the 2007-9 decline. I don't often use log scale charts, as unless there are clear trendlines on the log scale chart I assume that the normal chart should be used. A great example of where log scale is best however is for the 1974-2007 secular bull market. This formed a perfect rising channel and in 2007-9 a little under 61.8% of that move was retraced, with the three key lows in 1997, 2002 and 2009 forming a perfect support trendline before the new bull market began from the 2009 low.

Again the 2008-9 period was pretty wild, but the target was in the obvious range and makes perfect sense on the chart as the secular bear market retracement of this long secular bull market.

2007-9 Crash - 1980-2015 SPX Monthly Chart:
So what's my point here? Well I have two points to make here. The first is that, however bad it may get in any bear market, ultimately the bulls always win, it's just a matter of finding a low and then they slowly (1987) or quickly (2009) get back into the saddle and stay there for quite a while. The best trade that you can make in a bear market is to go heavily long near the low, and stay that way for a long time.

The second point is that having charted SPX in detail back to when data for it was first collected in the early 1920s, the only decline in that time that doesn't like some kind of fibonacci retracement is the 1929 crash, where the real economy crashed with the markets due to misguided policies such as protectionism and allowing a general collapse of credit, triggering The Great Depression. Whatever happens here, the larger risk to the real economy seems less likely to be too little debt than too much. This crisis too should pass, and as long as governments aren't complete fools, enough care should be taken to minimise the damage to the real economy during this crisis.

So what would be the obvious retracement targets this time?

Well the first two big support levels were rising support from the 2011 low in the 2500 area and then rising support from the 2009 low in the 2350 area, and both of those have now been broken. That eliminates the obvious support trendline options and opens up the standard retracement targets and key support levels below. The first of the standard retracement level options is the 50% retracement of the rising megaphone from the 2011 low in the 2225 area, and that was tested yesterday, but likely won't hold for long.

The next big area is a range between the 50% retracement of the move from the 2009 low in the 2020 area, and the 61.8% retracement of the move from the 2011 low in the 1955 area. If reached that would be a very likely area to see a strong bounce but, having come so close to that so fast, it again may well not hold. If that breaks then that opens the final series of targets, one of which will very likely hold unless governments do manage to trigger another event like The Great Depression, which is a possibility, but at the moment seems unlikely.

The last three levels to watch, one of which will likely hold as this bear market low, are as follows:

The first level is just above 1800, the level that Goldman Sachs was talking about in the early stages of this crisis while the SPX was still grinding to higher highs, and was the low for three significant monthly candles, in April 2014 at 1814, in October 2014 at 1820, and February 2016 at 1812. That is a well established floor and might well hold again if reached.

The second target, on a break below 1800, would be the 61.8% retracement of the move up from the 2009 low in the 1700 area. If seen, this target really should hold, so the third target would just be a backstop target, though it is a very strong level that I would very much expect to hold in the unlikely event that it is reached, and that would be a backtest of broken resistance at the 2000 and 2007 highs at 1553 and 1576 respectively. That's the blue shaded level on the chart below.

One of these last three targets will likely hold as this bear market low and, as I mentioned, once we see that low, the bulls should be back in the saddle for the next few years, so that will be a low to buy, and I'll be watching for it carefully.

SPX Monthly Chart:
So what do we need to be paying attention to on the bigger picture as this crisis unfolds? Well there is a careful line to be walked here by governments. The level of coronavirus infections needs to be minimised if possible, and if that is not possible, then ongoing infections need to be managed (flattening the curve) at a level at which health systems are not overwhelmed, which is the point at which the death rates spikes from bad (maybe 2-3%) to really bad (possibly 10%+). At the same time the economic damage from social distancing and isolation has to be carefully managed to minimise the number of businesses that are closing down now temporarily, but fail to reopen later, having folded due to decreased turnover without sufficiently reduced costs.

I'm not a fan of the way the US and some other governments have responded to this crisis so far, as they were unaccountably very slow on the uptake that there was a problem at all, but it is the management of the economic impact here and now that could mean the difference between a modest economic contraction this year, and maybe next year, and a potential depression that could seriously damage economic prospects for years to come, possibly also forcing the collapse of the massive house of cards of worldwide debt that foolish politicians and irresponsible central bankers have allowed to build up over recent decades. Fingers crossed that they manage not to drop this particular ball, but they are aware of this danger and seem determined to avoid it. The odds are that the economic impact will be limited and temporary, though likely at a cost of considerably increased government debt and very possibly persistent inflation for years to come.

In the event that a larger economic contraction was to be triggered, then that could open lower targets on SPX. I'd note that the 1936-42 bear market delivered slightly over a 76.4% fibonacci retracement, and the 1929-32 bear market managed to deliver an absolute decline of 86%. It is possible to go lower than the targets I have given above in really extreme circumstances, and I am assuming that governments will act to prevent those circumstances. I'm expecting to be right about that but I could of course be wrong. We'll see.

This has been a very interesting and surprisingly predictable market to trade and we have been delivering a lot of impressive moves for subscribers. If you'd like to see some of this analysis before the market moves, you can get free access to that for a month just by signing up for a 30 day free trial at theartofchart.net on this page here. That gives you access to most of our analysis and particularly our subscriber only twitter feed where we post our intraday updates and targets. The trial is free for 30 days, at the end of which you can either cancel with no obligation or subscribe to one of our services.

Friday, 20 March 2020

A Short History Of Superflu Pandemics

My last few posts have been a coronavirus COVID-19 series, so I'm putting in the links here so as to refer back to them easily for now. These are the six posts so far.

18th February - Peering Through The Fog Around Coronavirus COVID-19
24th February - Some Genuine Coronavirus Numbers Coming Through
28th February - Falling Down The Steps
9th March - A Tale Of Two Cities
12th March - Sudden Death
16th March - Pinball Markets

I was saying in my post on 24th February that if COVID-19 broke out to become a global pandemic then it would perhaps be the most predictable natural disaster in human history and this post is to explain why that is, and also the look at in detail how this is different from the normal surge in influenza cases that is seen in the winter months every year.

The image below was taken during the recent COVID-19 decontamination of Dolmabahce Palace in Istanbul and I'm posting it just because it is a beautiful image, and seems appropriate to use in a historical look at this kind of pandemic:


First though, I'll have a look at the stats we have seen in recent weeks on the spread of COVID-19 worldwide, and compare them to the numbers I was looking at in my post on 18th February. On 18th February there were 73,451 confirmed cases and, of those, there were about 1100 cases outside China. That was 30 days ago, and in the meantime cases outside China cases have risen to 154,546. Over that time therefore, cases outside China have been doubling approximately every four days and, just to put that in perspective, if cases were to increase at that same rate for another thirty days, then confirmed cases outside China at that point would exceed 20 million. To put that number in perspective it is estimated that in the 1889 pandemic, the 'Russian Flu', about 40% of the world population was infected, and in 1918, the 'Spanish Flu', about a third of the world population was infected.

In terms of the fatality rate, in terms of reported deaths per confirmed case that was about 2.5% on 18th February, and that has risen to about 4.15% now. The fatality rate in terms of reported deaths versus reported deaths per reported recovery that has dropped a bit from the 14% on 18th February to about 11.52% now. Overall the true eventual fatality rate may be between those numbers, I was suggesting 5% to 10% range on 18th February and that still looks reasonable, lower than SARS at 10% and a lot less than the more virulent Mediterranean version MERS at about 35%, though all of those numbers would likely be higher than the real numbers, as they would exclude many cases never picked up because those infected were asymptomatic or developed only mild symptoms.

On to the history.

The symptoms of human influenza were first described by Hippocrates about 2400 years ago, so flu in various forms has been with us for much of, and possibly all, of recorded history. In that time there have been many flu pandemics, though the further back that was, the harder it is to distinguish these from other disease pandemics, due to the lack of detail in the older historical records. In the first instance most of these pandemics were easy to confuse with the regular flu casualties in the winter months, but the difference was that the virus was a new strain to which humans had not yet developed immunity. After the first wave or three that immunity would develop in human populations and the virus would just become another of the many flu pandemic has-beens that return in the winters. Any new strain starts small, and unless you are familiar with the history of these pandemics, appears unthreatening in the early stages:

The fact that a new strain of influenza is new is the key difference, as exposure over time builds both herd and individual immunities, reducing the impact of subsequent outbreaks. The key example to illustrate that would be when Columbus 'discovered' the Americas. I say 'discovered' as the existence of the Americas was no surprise to the many people who already lived there, but the important point here is that at that point two human populations that had been isolated from each other for a long period, during which both had been exposed to, and built immunity to, those new diseases that had emerged during that period of isolation, introduced those diseases to another population that had not previously been exposed to them. The impact was dramatic.

In Hispaniola (modern Dominican Republic and Haiti), the Taino people who encountered the Columbus expedition in 1492 had a population of 250,000 at that time. By 1517 only 14,000 remained. The impact of the new european diseases on the native american population killed by some estimates perhaps as much as 90% of that population in the decades after they were exposed to them.

That works both ways of course, though the more populous europeans, also exposed historically to diseases originating in Asia and Africa, brought most of the new diseases to this exchange. It is argued by many historians that in return syphilis was introduced to Europe, initially in a very virulent form where the flesh would melt off the body, and death would be within week or months, rather than the decades that the later and milder forms of the disease took. This was a disease that had a huge impact on Europe over subsequent centuries, though not comparable in scale of course to the devastating impact of the diseases introduced to the american populations.

The point is that a new disease, or new strain of an existing disease, has a disproportionate impact as it first sweeps though the population, and every so often, as a sufficiently virulent new influenza variant develops and makes the step from animals or birds to humans, then it develops into a pandemic that can have a major impact, depending on the communicability of the variant and the parts of the population most vulnerable to it.

Over the last few centuries the first generally agreed influenza pandemic started in Asia in 1580, another started in Russia in 1729, another two started in China in 1781 and then 1830-3 and then the first 'modern' pandemic started in Russia in 1889, and for this and subsequent epidemics we have reasonably detailed numbers. There will most likely have been numerous other pandemics not listed before 'modern' times but only those with higher death rates will have made it into the records.

In this 'modern' period there have been five previous superflu (new variant) pandemics that infected at minimum 250 million people, and these were the Russian Flu in 1889-90, the Spanish Flu (originating region unknown but unlikely to be Spain) in 1918-20, the Asian Flu in 1957-8, the Hong Kong Flu in 1968-9, and the 2009-10 Flu (mixed american and asian variants). Four of these pandemics, all except the Spanish Flu, had relatively low death rates in the 0.03% (2009-10) to less than 0.3% (Russian, Asian and Hong Kong Flus), but the Spanish Flu had a much larger impact as the death rate was at least 2%, killing somewhere between 20 to 100 million people worldwide, out of a global population of about 1.8 billion at the time.

The Spanish Flu also had a disproportionate impact as unlike most superflu pandemics, which preferentially killed higher proportions of the very young, the old and the weak, the Spanish Flu epidemic preferentially killed the young and strong, as they were killed by their stronger immune systems, triggering something called a Cytokine Storm where an immune overreaction killed the virus hosts.


I've been reading about these epidemics since I was a teenager, as I've always been interested in history and read widely, so I've been watching new possible pandemic candidates as they've emerged and petered out over recent decades, such as SARS and MERS, or widened into pandemics with low death rates such as the 2009 epidemic. When I was writing on 4th February about the coronavirus post that I was going to write I was therefore already alarmed about this new variant, and astounded by the obvious lack of preparation that governments had in place for containing any new superflu pandemic, as it has just been always been a matter of time before the new superflu pandemic would emerge, likely from China, and every so often one of those pandemics would be virulent enough to kill millions of people. Governments should have been aware of this risk and at least somewhat prepared for this eventuality, and I remain astonished that this has caught them so unprepared.

So what about COVID-19? Well the fatality rate is still a work in progress, but it already looks at least comparable to the Spanish Flu from 1918-20. In terms of the sections of the population most vulnerable to the disease the numbers so far suggest strongly that the very young are among the least likely to die, which is unusual, but that the most likely to die are the most predictable groups, the old, the ill, the obese and the smokers, and that among these groups the death rates may be over 10%, though these numbers likely exclude numerous cases that weren't identified because there were no symptoms.  Western world populations have the oldest and most obese populations and that seems likely to be the reason that while Italy only has half the confirmed cases that China has admitted at 41k versus China's 81k, the number in deaths is now higher than in China. I would add a caveat here though in that while China may well have now stopped the progress of the pandemic, the quality of their numbers is nonetheless doubtful. Those particularly at highest risk are those in their 80s or 90s, and those with heart disease or diabetes.

So what is the overall health impact likely to be? Well the number infected has yet to be determined, and despite the very slow start that Western governments made in containing this pandemic, it may well be that the measures being taken mean that the proportion of the world population infected is considerably lower than the third of the world population infected in 1918-20. The outbreaks in Singapore and Hong Kong particularly seem to have been contained by swift and firm action and while they may not be typical in terms of their situations as islands with dense populations, that success may well be replicable elsewhere. The fatality rate might look higher on the numbers we have so far, but factoring in the large number of those infected (perhaps 50%) that seem to be asymptomatic or just develop mild symptoms, might be equivalent or lower than in 1918-20, with the deaths being in parts of the population that are less economically significant than in 1918-20.

The economic impact is another question of course. Much of China seems to have been shut down for at least a couple of months, and that is now happening in much of the rest of the world. This is reducing economic activity, reducing consumer spending and disrupting supply chains. This may well result in a world recession this year and, depending on the number of waves in this pandemic (there were three main waves in 1918-20), this may well reduce world economic activity for the next couple of years.


What about after that? Well historically periods where economic activity is significantly depressed by natural or human caused (notably war) disasters, tend to be followed by reconstructive periods of strong activity. The first world war and then the Spanish Flu developed into the roaring 1920s, and the second world war delivered an economic boom that lasted into the 1960s, with that boom being strongest and most durable in those most beaten down in the second world war such as Japan and Germany. The odds favor the period after this pandemic being a strong recovery, possibly considerably stronger than the generally tepid growth in much of the developed world since the financial crisis in 2007/8. We'll see.

I'm planning my next post over the weekend or on Monday,  and that will be the bigger picture levels and targets for SPX on this move now that the major support at rising support from the 2009 low in the 2350 has been broken.

A couple of other announcements to make:

Stan and I are doing a free public Big Five and Key Sectors webinar tonight at theartofchart.net at the RTH close at 4pm EST. That's covering AAPL, AZN, FB, GOOG, NFLX, TSLA, IBB, IYR XLE, XLF, XLK, XRT. If you'd like to attend then you can sign up for that here.

We have also scheduled a special free public webinar at 4PM EDT Sunday 29th March to look at COVID-19 and the Financial Markets, and in that webinar we will be considering COVID-19's impact on the economy, systemic risk and impact on dealers, the Fed's actions and credit risk, and reviewing potential trading opportunities. Be There Or Be Unaware! If you'd like to attend you can sign up for that here.

This has been a very interesting and surprisingly predictable market to trade and we have been delivering a lot of impressive moves for subscribers. If you'd like to see some of this analysis before the market moves, you can get free access to that for a month just by signing up for a 30 day free trial at theartofchart.net on this page here. That gives you access to most of our analysis and particularly our subscriber only twitter feed where we post our intraday updates and targets. The trial is free for 30 days, at the end of which you can either cancel with no obligation or subscribe to one of our services.