10 Month 20 Month Crossover in SPY

h/t to Larry Ledwick in his comment here:

https://chiefio.wordpress.com/2016/02/15/gold-vs-silver-vs-us-stocks-vs-eu-and-emerging-markets/#comment-67006

Here is an article which is flagging the approach soon of a cross over in the 10 and 20 month moving averages on the S&P 500 as a signal of the formal shift to a bear market.

http://www.marketwatch.com/story/the-sp-500-is-close-to-flashing-a-bearish-signal-it-hasnt-displayed-since-2008-2016-02-29

Rarely these days does something “new to me” come along on stock / investment timing or evaluation. I literally have a few dozen feet of books on the shelf on money and investing and I’ve read them all. For a while I was reading just about anything printed in the area. I’ve watched the “financial shows” fairly religiously since at least the ’70s. (Actually started with William Buckley on the local “educational TV” channel in the ’60s… but that was more Political Economy). To say I’m familiar with things like the 200 / 50 day moving average crossover and “deaths cross” et. al. is an understatement.

But I’d never bothered to look at the “10 month” scale of things. Since there are 5 work days in a week, and 4.3 weeks in a month, that ought to be about 215 days duration. But that’s the short end, the long end is up at 430 days. Just too long for my attention span.

Yet it looked like something worth a look. So I did. It’s very interesting.

Here’s the SPY S&P 500 ticker on a “all data” graph with monthly tick marks and the 10 month / 20 month SMA lines on it.

SPY all data 10 Month 20 crossover 3/2016

SPY all data 10 Month 20 crossover 3/2016

Click on it to embiggen (and then click on THAT image if it is still too small and it will get even bigger ;-)

First off, note that this is only 15 years of data. Way more than needed for a daily tick mark graph, but on this scale, it doesn’t give a lot of sample space to assure reliability. There are only 4 total crossovers, 2 up and 2 down. OTOH, it is pretty much ‘spot on’ for those events. What isn’t known is if this indication scale can have “sideways wobbles” or sporadic “dip and recover” or any of the other little bits seen on faster time scales. In short, how forceful are those crossovers and are they always like that, or just this sample?

So just hold that in mind and season any conclusions with that doubt as to reliability / resistance to shorter scale wobbles.

OK, looking at the graph, we see price fairly reliably tracking the 10 month average line on the uptrends, A “buy when it touches sell when gap above” on uptrends ought to be a good trade timer. On the downtrends, price tends to track the 10 month, but pulls far away at bottoms. You can clearly see price bar “failure to advance” and ‘double tops’ on the top inflections (and the ‘go flat’ of the 10 week SMA line). Bottoms are mixed. One is ‘double bottom’ and the other is a giant V spike. Less clearly defining bottoms, and the crossover off a bottom is a bit late for making a bundle. Still, the ‘top calls’ are reasonably fast (as tops ‘roll over’ more slowly than bottom spikes reverse – those short covers spike it up fast).

WAY over at the far right, you can barely see that the 10 month 20 month SMA lines have had crossover to the downside. IF this theory is correct, it’s saying we have a bear market (and they tend to last about 1 to 2 years).

I find it interesting that at this time scale, the tops have “kangaroo tails” (the thin price excursion line out of an open / close price bar) that point down. This implies that even at tops, the main price excursions to the downside happen mid-month with market makers opening / closing the month at higher prices. Interesting “Dig Here!”… Do prices have a monthly ‘return to the mean’ pressure at start / end of month? IF so, is it market makers, or monthly investment influx or?… (so buy mid month, sell month end to trade that?) For other time scales, the “kangaroo tail” tends to point away from the direction of momentum, so at tops you tend to get the ‘tails’ pointing up; at bottoms, down.

We do clearly see more red at the recent ‘topping action’ end of things (the 2015 year) and that makes the top more visible too. On rising bull market runs, the red months clearly show up as a ‘buy the dip’ moment (does it do that fast enough to be useful? If one buys the start of the next month, likely yes, given that ‘end of month start of month’ thesis on return to the mean noted above, but given the tiny ‘tails’ on those price bars, mid month of down months is likely the optimum timing. Another “Dig Here!”… best buy indicate inside this trend guidance time scale…

The only help I see in the bear market segments is that it says to stay clear of them. There is a “maybe” on the ‘tails’ of the down months saying “maybe a mid month buy” on bloody crash days and rapid sell on rise would be a decent day trade / swing trade, but trying to time ‘long in a bear market’ is a bad idea. Perhaps a “short and ‘cover on the crash days’ then re-short”? Yes, yet another “Dig Here!”…

On to the indicators.

RSI does a nice job of saying “touched 80 top soon” (where “soon” on this time scale is about a year) and “touched 20 that’s a bottom” with only a few months lag to show. Nice.

MACD topping crossovers are reliable and timely. The bottoms are also reliable, but a bit slow to ‘catch the bottom’. Mid-run up MACD xover is a false indicator as it is mostly showing ‘steady rate up’ and not an inflection of price. (price not accelerating, but continuing up trend). Coupled with the RSI top / bottom calls, though, it’s a nice enough confirmation signal. MACD Histogram (those black tick marks on the zero line) do a good job of marking bear vs bull markets. During prolonged runs up they tend to ‘go flat’ as it’s just a steady market with little acceleration; but with the red/blue lines clearly way above zero that’s still a nice ‘steady up’ indicate.

DMI / ADX is a bit hard to read at this time scale. “Red on top” (DMI -) still works as a ‘get out’ call but the inflections of ADX (the ‘strength’ line) are harder to interpret. During more sustained up runs the lines just wiggle together. Likely due to the ‘crash markets’ moving the range to very wide to encompass that extreme volatility. Still, just knowing “Red on top be out” is worth a fair amount for setting bear market context. DMI – went to “red on top” just about the same time I started saying this was a market top

Here’s a shorter time scale (5 years) with the price bars removed, just the 10 / 20 month SMA lines so you can see the recent crossover better:

10 Month 20 Crossover with Volume ROC and Momentum 3-2016

10 Month 20 Crossover with Volume ROC and Momentum 3-2016

Again, click to embiggen.

Now you can fairly cleanly see the recent crossover.

I’d also note that the 10 Month SMA line has a clear inflection to a ‘go flat’ at about the time I started saying ‘market top’ that is also consistent with the MACD crossover and histogram swap to below zero.

Looking at Volume, we can clearly see the downtrend in volume over the bull market as buyers get saturated (ever shorter black bars and the gold SMA line slope is downward). Then, at the topping point, more big red bars as price volatility sets in with more selling on down spikes. Note that the last month volume bar is very tiny due to the month just barely started. One wonders if the BigCharts method of graphing monthly data will be subject to ‘end effects’ and have the last data point flip back and forth over the month… A tiny “dig here…”.

I’ve also put ROC Rate Of Change and Momentum indicators on this graph. They are very similar in what they measure and how they show it. Sometimes one is a bit faster or more reliable than the other, depending on time scale and the particular stock. At this scale and for a broad index they are very similar. ROC is a bit ‘twitchier’ and went below zero in that 2011 ‘dip’ of significant size. MO keeps you aware it is STILL a bull market during that ‘big dip’. I’d likely use ROC as I’m more trade oriented, but a longer term perspective from this graph might benefit from the Big Mo saying “Steady there, it’s a dip in a bull run, buy it”.

Here’s a decade worth of data with daily tick marks for comparison:

SPY daily 3_2016 decade 66 Day vs 198 Day SMA

SPY daily 3_2016 decade 66 Day vs 198 Day SMA

This one REALLY needs that ‘click to embiggen’… though at least under Chrome it gives a list of the three graphs and you have to click the ‘give me the next one’ file name in the lower right corner under the graph to get to the 2nd and 3rd graphs. YMMV.

Due to BigCharts not letting me set specific days on SMA sets (it just doubles the first one for the second…) I’ve used the 66 day base (instead of 50) as that gives 132 and 198 day SMA lines in the “3” set. A 66 / 198 is going to act much like a 50 / 200, and only lag a few days on the crossover.

This scale does tend to have you ‘trade out’ in the major ‘corrections’ as in 2010 and 2011. Probably a very bad idea for long term investor types. I’m OK with being out of a few months on a trade basis, but having the longer term perspective of the 10 month 20 month would help with the courage to ‘buy the dip’ at the bottoms. That longer term does seem to better keep bear / bull market perspective.

The huge volume spikes at market bottoms shows up well, as does the long drift to lower volume at the end of a secular bull market run. Our recent volume spikes have been on down price swings. The market is a volume seeking machine, and it is finding more volume (so more commissions) to the downside right now.

MACD and DMI are just darned hard to read at this time scale on this long a duration chart. Still, you can make out that MACD is mostly above the zero line in bull runs, mostly below in bear markets, and that the crossover / inflection points are good trade moments. DMI red vs blue is generally helpful for trade context (buy the dip or sell the rip in good times, cover shorts or short again in bad) with timing based on the red or blue line inflecting.

Overall, for this long a time span, the daily data are just too volatile to be easy to read. The monthly data are making very clear statements about market status as a secular bull or bear market.

Usually at this time duration, I use the weekly tick mark graph. Here it is:

SPY weekly 20 SMA stack 3-2016

SPY weekly 20 SMA stack 3-2016

MACD clearly above / below zero easier to ‘read’ than the daily graph for context. DMI inflections and red / blue crossovers a bit more usable for slow (trend) trades. Probably a bit less useful for long term investment context compared with the monthly graph.

In Conclusion

I’m liking the idea of using that monthly tick mark 10 / 20 graph as setting my ‘market context’. It takes a lot less ‘interpretation’ and knowing when to ‘ignore that little wiggle’ than the other time scales.

ALL the graphs are presently saying the same thing: We’ve swapped over to a bear market and it is not a time to be ‘long stocks’. At tops and bottoms, I move to faster time scales and faster trades, so I’ll “go long” for a fast trade on a ‘return to the SMA stack from below’ (a very risky trade, BTW). These longer duration views are good to keep in mind then as a “nag” that “this is just a risky trade, not a bull market”. Similarly, in down weeks in a long bull run, they would remind to “Buy The Damn Dip!”.

As a final “Dig Here!” it would be interesting to apply this time scale to a bunch of other tickers and see what shows up. Perhaps bonds and commodities vs SPY / QQQQ / DIA / RUT? Does EEM Emerging Markets move before or after the US Markets? What about EZU Eurozone? (A ‘quick peek’ showed it about the same back in 2007, but never recovering as much as SPY and ‘breaking’ first in 2014, about a year ahead of the US markets, but driven by what?… Hmmm…. another “Dig Here!”)

So yet another tool, yet another context setting device. I think this one is very useful for keeping your heading in the long duration business cycle.

Subscribe to feed

About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in Economics - Trading - and Money and tagged , , . Bookmark the permalink.

14 Responses to 10 Month 20 Month Crossover in SPY

  1. philjourdan says:

    I agree about the bear market. But wonder if this is a post indicator that they sought out after the fact. As you indicate, there is not enough history here to see if it is an accurate indicator over the long haul. But it does bear (no pun intended) watching.

  2. Larry Ledwick says:

    I thought you might find that interesting! Your more in depth analysis and perspective helps me understand it better as well, thanks. Having a relatively clear and unambiguous signal makes it easier to establish the over all situation in the markets with out a lot of in depth digging. For someone like me who does not actively trade but wants to be intelligent about which funds to use in the 401K and just to understand the general economic direction of things is seems to be a good benchmark to work from.

  3. Larry Ledwick says:

    Given the typical rate of change in tops and bottoms this comment:

    Still, the ‘top calls’ are reasonably fast (as tops ‘roll over’ more slowly than bottom spikes reverse – those short covers spike it up fast).

    Makes me wonder if you need to use two different pairs for example the 10/20 for calling the top, and a faster pair like say 5/10 to call the bottom and to indicate it is time to jump back in.

    Might be worth running some comparisons to see how other intervals shake out.

  4. E.M.Smith says:

    I handle it by shifting to a faster chart at tops and bottoms. Typically 1 yr daily…

  5. Larry Ledwick says:

    Possible hint about why the perception is that we are not in a recession based on short term market indicators, but the general public is feeling under pressure. Perhaps it is because consumer spending is holding up fairly well but only because consumers are leaning harder on their credit cards.

    http://www.cnbc.com/2016/03/10/us-credit-card-debt-balloons-to-917b-what-it-means.html

  6. E.M.Smith says:

    A $ Trillion debt here, a $ Trillion debt there, pretty soon you are talking real debt…

    It works out to about $1000 per family added. Something to note, but not covering a missed paycheck or three…

  7. philjourdan says:

    I think you are off by a factor of 10. It is $10k per family (assuming 100 million families).

  8. E.M.Smith says:

    @Larry:

    The Fed about to raise rates and folks have levered up as many bond shorts as they can? I’m shocked, shocked I say, to find gambling going on in the casino!!
    /sarc;

    Hell, I’ve been waiting for a chance to short bond via the tbt, but the charts have not supported it…

    Now, with the EU doubling down on negative interest rates, that will increase the demand for US securities even more, extending the time to when the short is ok to go.

    IMHO, it mostly says the traders entered the short too soon, and got caught by “unprecidented” Central Bank rate cuts. Now they are scrambling to figure out WTF…

    Oh, and I’m feeling happier about not having entered the short yet… Trust The Graph, Luke!

    @PhilJourdan:

    Sorry, I was unclear… That was a swag (Scientific Wild Ass Guess) of the size of the ADDITION to c.c. debt, that is about 1/10 the total of almost $Trillion… but one had to click the article link to see that number, so my statement seems in the context of the $Trillion. My bad.

    The point I was intending was that the delta of $1k wasn’t replacing the paycheck for that delta time. Yes, folks are carrying about $10k, but the add isn’t a check replacer and can just be the Christmas bulge having a hangover. Bad, but not horrible.

  9. E.M.Smith says:

    Probably ought to point out that with China a mess driving a flight to safety, and the EU hard negative rates and doubling down, that will drive more $ Strength and drive money into the US bond as a higher rate alternative, the EU action will drive demand up for the US securities, right when the trade (absent the ECB EU effects) ought to be a short on US Treasuries; since a 0% bond is worth less when The Fed is issuing 1% bonds. Any trader not watching China and the ECB was caught out… ought to have paid more attention in their International Macro Econ class… (one of my favorites, BTW. Got a very good grade, but not sure if it was an A or just under. IIRC, it was an A, but that was a long time ago…)

  10. Larry Ledwick says:

    Speaking of China’s investors trying to buy properties off shore to preserve wealth.
    http://www.foxnews.com/travel/2016/03/14/starwood-receives-nearly-14b-buyout-bid-from-chinese-group/

  11. Larry Ledwick says:

    What happens when this last buyer gives up and either can’t or won’t buy any more?
    http://davidstockmanscontracorner.com/theres-only-one-buyer-keeping-sp-500s-bull-market-alive-c-suite-stock-option-collectors/

  12. E.M.Smith says:

    Interesting point… but the article seems to think buybacks are only fueled by rising profits. Often they are used when profits are NOT rising as a way to improve the “earnings / share” by making shares outstanding smaller… With interest near zero, issuing a bunch of bonds to buy back stock lets you ‘juice’ some of the per share earnings numbers when real sales and profits are flattening.

    Still a bad end game, but with a few more wrinkles along the way…

Comments are closed.