Macro Cycles 2014-2018

[shouldaindex]

The last decline to reach the bottom of the last 8 cycles has lasted between 2 to 8 weeks.

If 4936 turns out to be the bottom (I'm unsure at this stage), it's last decline lasted 3 weeks, so it does fit within that pattern.

We are currently on the up, so if we are to make a new last decline, this information suggets it's likely to take another 2 more weeks to form a last decline (starting from the moment the upwards trend turns).

 

[shouldaindex]

Bottom Picking:

This one's very roughly (not exact figures) from reading XAO graphs tonight, but the overall information still paints a picture:

These are the very last drops to a find a cyclical bottom:

2011 7 days, -9%
2007: 5 days, -7%
2002: 9 days, -11%
1998: 13 days, -6%
1994: 15 days, -3%
1992: 8 days, -6%
1989: 9 days, -5%
1987: 3 days, -8%

These are the very first rises after a cyclical bottom:

2011: 11 days, +12%
2007: 6 days, +12%
2002: 6 days, + 6%
1998: 16 days, +11%
1994: 10 days, +5%
1992: 6 days, +7%
1989: 4 days, + 5%
1987: 4 days, +10%

So using 2011 as an example. If all my analysis (hypothetically somehow) points me towards an 18 day window that will have a multi-year bottom, I could still miss out on 9% by being 7 days too early, or 12% by being 11 days too late.

That's a tough business.

 

[rimtas]

I've been thinking about the nature of catalysts for a bear market..

Have you considered an idea, that there is no need for catalyst, neither for bear, nor for bull. If social mood goes down, that is affecting how people behave-they spend less,, save more, avoid debt, don't make far reaching optimistic decisions. This translates into lower earnings, declining investment and lower share and realestate market.
Because share market is the best known indicator of social mood, it predicts everything else, because you sell shares instatntly, but other decisions take time to materialize-thats why earnings lags stock market, as an example. You can not predict share market looking at earnings, but you can predict earnings looking at share market.


Also, your cyclycal calculations are based on linear projection-you try to use past few decades corections and assume market will behave the same. You get sometimes confused that market once ot twice produced "out of normal" corrections, when in fact they are normal, if you consider that market is not linear, but behaves as a fractal-big corrections occurs on 1 min, 5 min, 1H, I day, 1 Week, 1Month and 1 Century intervals regularly, they are normal.

I must admit that Australia's market is the Only in the world that in the last 150 years produced only short lasting(~5 years) and least declining (bigest drops were ~50%)corrections. It's entire chart fits nicely into two parallel lines. But this nice linear statistics ends right here.
From 2007, All Ords are still below ATH, meaning that 5 year correction average has bean broken-it is already almost 8 years below ATH and heading down. First time in 150 years. This should tell you something. It tells me clearly-we are heading for this, as you consider-"not normal, not average" correction.

 

[shouldaindex]

Hi Rimtas, thanks for your thoughts on both Earnings and Historical Information.

I definately don't stick literally to the information/analysis as it has obvious limitations, but thought I'd jot it down as an exercise. It's one tool amongst infinate.

 

[db94]

Hi Rimtas, thanks for your thoughts on both Earnings and Historical Information.

I definately don't stick literally to the information/analysis as it has obvious limitations, but thought I'd jot it down as an exercise. It's one tool amongst infinate.

While Rimtas may be raining on your parade, I'm loving this!! Really insightful. And putting some things into perspective for me. Keep up the good work

 

[shouldaindex]

Thanks DB for your feedback!

Here's a portion of the ASX historical I've used before, and I'm going to look at secular bear markets and what happens in them. The past 9 cycles in the past 35 years have been in a secular bull market, so let's present an alternative to the future.

Secular Bear Markets
(graphic isn't great, but read below)



Worst Long-Term Bear Markets:
1929 to 1945: 16 years / Annual average gain of about 1% (plus dividends)
1967 to 1983: 16 years / Annual average gain of about 2% (plus dividends)

If we applied that secular bear scenario to 2007, where the ASX gained between 1 to 2% annually, the ASX would be 8000 to 9400 in 2023.

The interesting thing is that would still still be a healthy gain of 6-8% (plus dividends) in the next 8 years (5000 to 8000-9400 in 2023)

So we would have to create a new record worst history which would make all past information irrelevent! Otherwise if it follows the past within a certain range that is an approximate guide to what may happen even if we are in a 16 year secular bear market.

 

[shouldaindex]

GDP growth for Q2 = 0.2%

All it would take is a further -0.3% drop to get into negative territory for 2 Quarters and that's a recession.

If a major economic event did happen to end this cycle, we would then be looking at a more extreme set of data that includes Recessions/GFC/Crashes. That would bring into play the possibility of a 30-60% decline, where as a cycle without such an event has historically maxed out at about 22%.

Data since 1960:

Cycle ends with major economic events (rounded figure %) 20, 30, 50, 20, 30, 50, 30, 50.

...Without (rounded figure %) 20, 15, 20, 20, 15, 20, 20, 20.

Up until today, I hadn't read any data to indicate any danger of this, but it has upped my alert to needing to not be so locked into a cycle end without one.

 

[shouldaindex]

This is why I do this for a hobby.

Using Intraday Highs and Lows I've managed to turn the 15% declines (which were rounded from 17%, 18%) in 1998 and 1984 into 20% declines (using intraday highs/lows instead of just closing), which means every mention of 15% now refers to 20%.

So now that cycles have all ended with 20%+ declines, for a 30+ year pattern to continue the ASX would be expected to get to at least 4771 (20% drop from 5963).

 

[shouldaindex]

When the ASX dropped 20% to end it's cycle (over the past 30 years), the specific intraday peak/trough ranged from 20.0% to 22.9%.

So if that happens in this cycle, the range we are looking for are: 4599 to 4771 (-20.0% to -22.9)

On a similar gradient from April to August, gets us to that range by October.

Again, not meant to be taken literally, but gives you some sort of expectation.

 

[tech/a]

Wow.

Rational randomness.
Well one interpretation.

 

[shouldaindex]

Found this regarding Fed Rate Hike Cycles



Seems like PE contraction does happen, whether the 10-15% contraction of the last few months is the contraction as part of this phase of the cycle, or if there will be the usual post-hike contraction, we'll wait and see.

But historically it shows the 3 months after are the worst period during the 12 month period surrounding the first hike. Also, it looks like we may set a record loss for the 3-6 months prior, so it shows the limitations of historical patterns.

 

[shouldaindex]

One of the discussions regarding using 30-40 year sample sizes, is that it may actually be capturing an abnormal period of history over the past 100 years.

So what I've done is drawn a trend line from 1915 to 2009, which I consider to be 'fair'.

ASX 1915 to 2009 Trend Line



I've calculated that, that trend grows at 5.6% Per Annum.

Extrapolating that to 2016 would have us at the 100 year trend @ 4500.

Keep in mind that very small % changes have a drastic effect over 100 years, so take it with a grain of salt. What it can illustrate is that cycles can be at different stages within larger cycles or what might be considered trend within one time frame may be considered above or below. And also just like a sample size of 30 years, a 100 year sample size may not continue in the future.

 

[shouldaindex]

Wanted to touch on the current cycle in the context of the larger cycle using the Dow Jones (chart sourced from a google search).

Dow Jones Historical



Look at the blue boxes that lay across 17 year bear markets between:

1905-1922
1932-1949
1965-1982
1999-2016 (projected)

You can see similar traits that have occured in the current 17 year period such as tops matching or gradually getting slightly higher (tick), biggest crash in the middle (tick), and smallest decline at the end (hopefully).

The good news is if we are at the end of a secular bear, the next period is where the market makes gains that it doesn't retrace in a secular bull.

Like everything there is contrary information and arguments (The 5% trend line has us at about double the expected price), so this is just one piece of the puzzle.

 

[shouldaindex]

When the All Ordinaries hits exactly a 20% decline from peak, where is it 12 months later?

Date: 12 Months Later %
Sep 21 1990: +9%
Nov 16 1992: +53%
Nov 22 1994: +14%
Oct 28 1997: +16%
March 5 2003: +23%
June 11 2008: -26%
August 8 2011: +6%

So I've mentioned splitting up forecasts based on WITH major economic event and WITHOUT. I've previously categorised 1990 and 2008 as WITH due to Recession and GFC.

12 Month Returns once ASX hits exactly -20%:
WITH Major Economic Event: +9%, -26% = Average -9%
WITHOUT Major Economic Event: +53%, +23%, +16%, 14%, +6% = Average +22%

Doing this for this cycle, our 20% decline number is 4771. So if you had a strategy of investing that day, this is the types of returns that have occured in the next 12 months.

 

[Junior]

Shouldaindex, really enjoying the data in this thread. Keep it up.

How is an 'economic event' classified? Is it likely that 2015 or 2016 will end up in this category?

We've had a commodity price and oil price meltdown, and Australia likely to be in recession. Do any of these factors indicate an event?

 

[shouldaindex]

Shouldaindex, really enjoying the data in this thread. Keep it up.

How is an 'economic event' classified? Is it likely that 2015 or 2016 will end up in this category?

We've had a commodity price and oil price meltdown, and Australia likely to be in recession. Do any of these factors indicate an event?

I'd define it as being 100% obvious. That way there's less room for 'fixing' subjectively.

IE. Stock Market Crash in 1987, Recession in 1990, GFC in 2008.

I'm less sure about analysing the actual economy than the stock market, that's why I tend to seperate things into the 2 categories "With or Without A..."

But out of every piece of information I've posted the one that sticks out to me (globally, that will also reflect in Australia) is how long U.S Earnings have gone without a 10% decline. We're at 8 years, and there's only really 2-3 other periods that have gone longer since 1880. This is mirrored by the 3rd longest bull market in the U.S.

https://www.aussiestockforums.com/forums/attachment.php?attachmentid=64083&d=1440756106

Also the House Price cycle, there's never been more than 5 years without a decline and we're in the 4th year.

http://www.rba.gov.au/speeches/2008/images/sp-so-270308-graph1.gif

So if the ASX gets into my target range, I'll probably have to split my buys into tranches because I'll want a spare bullet if things do reach the major economic event stage.

 

[Smurf1976]

How is an 'economic event' classified? Is it likely that 2015 or 2016 will end up in this category?

We've had a commodity price and oil price meltdown, and Australia likely to be in recession. Do any of these factors indicate an event?

The price of oil has dropped about 60% compared to a year ago. Given that other commodities have also fallen, I'd say that's an "event" at least in that it's a very significant change in the price of a key commodity. It's not as though we're talking about a tiny market or a small change, the change in the price of oil is roughly a $2.4 Trillion annual change (in USD) at the global level.

Where it gets more complex is that the "event" in this case is potentially a good one for many countries and most consumers. Australia, the USA, Japan and China are all net importers of oil (yes, the US imports - production there has risen as is well known but they're still an importer overall).

 

[shouldaindex]

S&P500 15%+ declines since 1929. I've grouped them by (rounded number)

(85) - 83
(55) - 54, 54, 56, 56
(50) - 48, 48, 49, 49
(45) - 44
(40) - 40
(35) - 33, 33, 34, 36, 36
(30) - 29, 29, 31
(25) - 26, 26, 26, 26, 26, 27, 27
(20) - 19, 19, 19, 19, 20, 20, 20, 20, 22, 22
(15) - 16,16,17

So playing around with a few takeaways / assumptions.
- 19%+ declines happen approximately every 2.3 years. We are well past that in 2015.
- Only 3 times out of 40 not reached 19% if reached 15%. S&P500 triggered the 15% mark last month, so statistically there's a 93% chance of going lower.
- I would feel comfortable predicting a decline of under 30%, without a major economic event such as U.S Recession or Financial Crisis.
- Using probabilities to gain an extra couple %, you can see where the groupings have settled most commonly around round numbers.

S&P500 Years in between 19% declines since 1940
6, 2, 9, 5, 4, 2, 3, 4, 3, 5, 6, 3, 10, 7, 4, 4

We're in the 2015 Bear phase of the cycle now, and I'm tentatively expecting another in 2017/2018 due to some other indicators, meaning a shorter than average bull phase of the next cycle.

 

[rimtas]

You will have almost the same results if you try to do this analysis on 5 min , or 1H charts(not percentage of course). You will have plenty of data in smaller timeframes (not just a few centuries)and notice that smaller declines occurs more frequently than bigger ones, and crashes also are the norm, just they are the rarest. It is a no brainer to see that the same picture occurs at any time frame.

 

[shouldaindex]

ASX All-Time Bulls and Bears



1900-2007:

Bears - Blue retraces and goes nowhere - 57 Years
00-02
11-16
22-30
33-42
48-53
59-74
79-82
86-92
03-07

Bulls - Orange makes permanent gains - 50 Years
02-11
16-22
30-33
42-48
53-59
74-79
82-86
92-03