I find this a very interesting chart. It’s a 10 year chart, but has the daily ‘tick marks’ on it. That makes it a bit ‘busy’, but with interesting information on it.
SPY is the S&P 500 in the USA. RUT is the Russel 2000 (so smaller stocks). EWG is a Germany Exchange Traded Fund while EWU is a United Kingdom ETF. TLT is a US longer duration bond fund, so ought to move in opposition to stocks most of the time.
I’ve put volume, volatility and MACD on it. Originally I had Momentum as the bottom indicator, but it wasn’t as interesting.
OK, first thing to realize is just now much Germany has high volatility compared to the UK and US markets. Golly… It can have quite a ‘bubble’ when things are bubbly, and can really crash when things are crashing. Even more volatile that the USA small cap RUT index.
Next, we notice that after the last crash, the UK did not recover nearly as well as Germany. Furthermore, it was a “RUT Match” going into the top, but didn’t even recover with the RUT coming out the other side. I presume this is due to some higher exposure to financial stocks and that it enjoyed them during the bubble, but not in the aftermath…
Now look at TLT vs SPY and vs EWU. Rather terribly uninteresting going into the bubble. Then it starts to have a clear alternating motion coming off the top of the bubble. Spiking up on any sell off / panic. Stocks rolled over A LOT off that top, before TLT started to rise. While since then, a 30% to 50% TLT component would dampen swings (and you would get a relatively stable rising line between TLT and SPY / EWU as your line of gain) that was NOT true prior to the bubble (so might not be true in the future either). Most likely this is an artifact of all The Fed intervention and both interest rates being pushed down to near zero and The Fed buying $Trillion traunches of bonds from time to time… But at some point The Fed will need to stop doing that.
Look at the last few of years. BOTH Bonds and Stocks rising, but with alternating ripples. I think that’s a clear indication of a Fed driven “bubble” building in US Bonds. Note, too, that in the last year the UK has decoupled from the SPY / Bonds movement. It would be better to compare EWU with UK Gilts, but thats a bit hard to do with US Centric stock tickers.
Now look at Germany in the last few years. It has failed to really make progress since the initial recovery from the crash. The second year it got a bit higher, but generally it’s just a ‘flat roller’. Still, that’s about a 25% range, once a year. (Though the frequency looks to be ‘picking up’; which often means a breakout ‘soon’ when it comes to a point.) If you look at RUT, it’s failed to advance past those prior peaks for three cycles now. BUT… It has ‘higher lows’. That argues for an eventual breakout to the upside. (Not always, but often. It tends to indicate buyers buying sooner on each dip. But eventually buyers can get tired of buying…) In any case, range is getting smaller and frequency faster. Eventually we reach a point of the pennant and then ‘something has to give’.
On this time scale, it’s a bit hard to read Volume and Volatility. Why? Because it’s SPY and NOT the actual S&P 500 volume. That ETF is a distinct instrument and this is the part of the S&P 500 sales that happened only in that one fund. The fund has grown in popularity over time, so the early years are basically too tiny to read patterns well. (Or at all). But we can still see patterns.
First off, and easiest to see, a massive volume and volatility spike in the big crash. Then similar spikes at other “buy points”. So you want to buy when a volume / volatility spike has happened, and is starting to diminish. Harder to pick out is that the time to SELL is when volume and volatility have gone soft. Look at Volume first. Notice you get “holes” under the moving average line of volume? It happens a bit before the drop. Notice that “lately” the volume and volatility are both rather low? It’s time to start worrying… (as if we haven’t been worrying all along ;-)
Now look at volatility. Look for patterns of ‘spikes up’ at buy bottoms, and low flat points near sell tops…
IMHO, using a product of volume and volatility as an indicator would clarify both. ( I’ve taken to calling this “VolVix” but have not automated calculating it, yet…) With a bit of practice, you can just ‘eyeball it’, so making it an automated calculation has been low on my list of ‘things to do’. It does ‘have some issues’. Volume drops at Holidays. There would need to be a ‘holiday’ weighting applied. It works best on aggregates, like indexes and whole markets, a bit less well but OK on index ETFs, and gets less usable as you move to more thinly traded things and individual stocks. I’m not seeing a ‘whole lotta luv’ in the VolVix at the right side of the graph…
At the bottom I’ve put MACD. In this case I’m looking at ways to read it that are a bit different from the usual ‘crossovers and plus / minus’. Notice that it makes relatively large ‘below zero’ excursions at those VolVix peaks (buy points) and that it is only really a good idea to own stocks when it is leaving one of those downward excursions for a bit of time in the above zero range. Picking an ‘exit point’ is much harder. Sometimes it drops TO the zero line, then goes back up. Especially in the early faster rise out of the crash. During the long ‘lead in’ it’s more of a simple oscillator, and after the major recovery, it’s been ever fewer ‘counts’ of wobbles above zero before the next big dip below. Frequency is picking up and approaching “1″ again. For a while post crash it was ‘three humps and a dump’, then weaker, and now it is more like ‘dump the hump’ on a crossover to the downside. My typical method of dealing with that is to use ADX to tell me to move to a ‘faster chart’ at times like now, but perhaps a longer term view, with a frequency scaling applied, would also help… Or just using ‘trade rules’ that buy puts when volatility goes low (so they are cheap) and then you use them if/when you know it’s a ‘sell time’…
At any rate, I find this chart informative.
Looking at the SMA lines, you will notice I’ve shifted them to 100 and 200 day. That’s a longer term time horizon. IMHO they make a very clear ‘in and out’ indication on crossovers, especially on the exit at the tops.
This next chart has EWG, EWU, and RUT removed (so the scale can expand and make it a bit easier to read for SPY). I’ve also shifted to a 1 week tick marking. (And the SMA lines are 20 and 40 weeks, of 5 days each, so ‘the same’). Now MACD is more ‘well behaved’ with ‘above zero’ being ‘hang in there’ and below zero being “be out until a buy spike indicate and / or crossover”. It’s a bit harder to see detail in the Volume and Volatility, but the big chunks stand out more clearly.
It looks to me like using these two, together, gives a fairly reliable ‘context’ for major buy / sell times. As MACD is still above zero now, it implies we are still in a positive run, but not at a ‘buy point’ spike down yet. I’ve also put PSAR on this chart. Those little red dots (click the image to get a much bigger easier to read one). They give a pretty good ‘exit’ a the tope, though careful use of stop loss orders might do as well.
Overall, the “shape” of the patterns argues for things getting tired. Shortening swings. Higher frequency of trade points. Narrowing of the advance range. Even the last ‘buy the dip’ having a relatively low volume / volatility spike. Things look more like ‘approaching a top’ than ‘buy!’. So while it’s still a ‘be in’ over all indicate for the SPY ticker, it’s not a very strong one and it’s not an ideal buy point. Furthermore, RUT was looking even more tired and the European tickers were looking worse. Yes, one ought to do a specific chart for each of them. It’s not hard, but would take a long time to ‘write up’. So you are encouraged to do your own as ‘homework’. Furthermore, there was talk of the BOJ activity likely making Japanese stocks interesting, so doing one for EWJ would be a good idea too. (Though since they are acting to weaken the Yen, what the BOJ gives to the Japan Stock Market the Yen may well remove…) The biggest “issue” with doing that, though, is that the individual non-US country ETFs often don’t have enough volume (or enough recently) to give a good volume indication / spike.
So “hit the link” for Bigcharts and look at the ones that particularly interest you and see if they have enough volume to ‘work’. I’d rather have an actual ‘ticker’ for the market volume itself, but Bigcharts doesn’t have that, near as I can tell.