Following the spike down into Monday morning's low we've seen quick short-covering rallies (inspired by overnight rallies in the futures market) but then sellers quickly took over. That almost happened again today but the selling has been decreasing in intensity and we could be looking at least a short-term bottom.
Today's Market Stats
Today started out like Tuesday with a large gap up following an overnight rally in the futures, which ostensibly was due to more "good" news about China pulling out the stops to get their stock market to rally. So far their efforts have failed miserably and in fact have had the opposite effect by scaring investors out of the market. But it had a positive effect on our market today as traders did some more buying following the pullback from this morning's gap up. A larger price pattern suggests we'll get at least another leg down before setting up a larger bounce correction and it's possible we'll see the market work its way lower over the next couple of weeks but it could be a very choppy affair but at the moment it's looking like we might get at least a slightly larger bounce (in time if not price) before heading back down.
These are wild times in the market as multiple-hundred point swings in the DOW create excitement and fear, all at the same time. We haven't seen price swings like this since 2008 and it does make you wonder if we've entered a similar phase for the market. Bulls will contend it was just a scary pullback into what will be just another v-bottom reversal. Bears argue the technical damage to the charts says this time it's different and is finally reflecting some reality in the global climate (a different kind of climate change). I'm siding with the bears here but as always, we'll let the price pattern develop a little more and make some judgments based on what price tells us, not what CNBC and others (your humble analyst-writer included) tell us.
The trading in the past week has highlighted a problem that has been plaguing the market for some time -- liquidity and its evaporation when it's needed most. A fellow trader, Bob, asked the following questions:
"In Monday's trading there were many large-cap stocks and ETFs that cratered between 9:29 and 9:31 (I watched QQQ trade at ~ $97 at 9:29 and then trade UNDER $85 at 9:30!!), which to me indicates that many, if not most, of the 'liquidity providers' pulled their bids faster than in the blink of an eye and created the resulting price 'vacuums.' How exactly can HFT [high frequency trader] shops claim, without laughing hysterically, that they provide liquidity when they can and do remove their bids, resulting in that 'liquidity' vaporizing in nanoseconds? Is it really liquidity if it's gone in nanoseconds and isn't there when it is needed most?"
To me this is more than an academic question because it highlights the problem with our current computer-traded market. The specialists and market makers are pretty much gone now, replaced by computers matching our orders. In front of our orders are the HFTs and they're typically in and out so quickly that they do provide liquidity for us "normal" traders. But when they step aside and take their orders with them it can present a sudden liquidity problem for the market.
As Bob highlighted above, are the HFTs really providing a service to the market or have they instead made it more dangerous? The answer is both -- during normal times the bid-ask spread is very narrow as orders are firing away on both sides. But when a big move happens the HFT algorithms take the computers off line until the market settles down (their split-penny trading suddenly goes to dimes and quarters and that's not how they trade). Poof, there goes the liquidity in a nanosecond and everyone else is left scrambling to find someone on the other side of their trade.
The "someone" who gets matched up with your order might have stuck an order in that's dollars away from the current price. He might have placed a buy order in months ago, figuring if SPY ever gets down to last October's low (near 182) he'd gladly be a buyer down there. But if decline in SPY hits your market stop order at 185 and liquidity goes poof at the same time, that 182 price could be the nearest matching price to your sell order. Hence the "vacuum" left below the last price as soon as the HFTs pull out. This is part of the explanation for the flash crashes we've seen and will likely see many more, some a lot worse than what we've seen so far.
There was an article in the Wall Street Journal and at Bloomberg Market Flaws that also discussed this problem with lack of liquidity, especially in ETFs. Could the cracks in the dam now finally be getting highlighted? What to do about it is the bigger problem. Shut down HFT houses? Demand market makers make a market at all costs? Perhaps the next real crisis will have the Fed announcing a program to back all banks who make a market (they've done similar things in the past when we've had market disconnects).
As mentioned in the article, there is now so much trading in ETFs, instead of stock picking, that when someone wants to sell (or many someones) they now sell ALL stocks at once. SPY is one of the most heavily traded ETFs and when one buys or sells the SPY they're effectively buying or selling all 500 stocks in the S&P 500. But that balancing between the ETF and the component stocks typically happens during the day whereas a sudden move often leaves the ETF disconnected from its component stocks. As reported in the article, "dozens of ETFs traded at sharp discounts to their net asset value (NAV) -- or their components' worth -- leading to outsize losses for investors who entered sell orders at the depth of the panic." Equity futures trading was halted shortly before the cash market opened Monday morning. The value of the VIX wasn't published until 10:00, by which point the DOW had already traded through the vacuum and lost nearly 1100 points. That's the kind of risk we see when liquidity vanishes in a nanosecond.
The article highlights a couple of ETFs and the general problem with circuit breakers hitting stocks and halting trading on them. The ETF market makers were unable to figure out the true value of the ETF and therefore started underpricing their sell orders and overpricing their buy orders so that they didn't take on too much risk. That's what created the low prices relative to their NAV and ETF sellers took it on the chin. Examples included the Vanguard Consumer Staples index ETF and the Vanguard Health Care index ETF both collapsing 32% in the opening minutes. During that time the value of the component stocks in the Consumer Staples ETF dropped only 9%.
All of this is only scaring more investors out of the market, which merely exacerbates the liquidity problem. This problem is not going to go away anytime soon and it sure is a caveat-emptor market. Keep this in mind when we're looking at the potential for a strong decline, what I often refer to as a disconnect to the downside.
Since Monday's low we've seen bounce attempts, helped by the overnight rallies. Today was the first day that sellers did not overwhelm buyers into the close, which could have bullish potential. The selling in the market could be close to drying up as margin selling finishes. Today's strong rally certainly helps in that regard. The market is showing short-term bullish divergences as the selling pressure wanes and tests of Monday's lows has it looking like we might be setting up for at least a bigger bounce. But a short-term pattern suggests we should get at least one more test of this week's lows, if not new lows, so it's not a time to get aggressive on the long side (or short side for that matter). If we get something stronger and impulsive to the upside I'll reevaluate what's happening but right now I would not trust the upside for anything more than a bounce. If we get the test/new low I would then be more inclined to try the long side.
Getting to the charts, with major uptrend lines broken when viewed with the log price scale, I've switched to the arithmetic price scale on the SPX weekly chart below to show where traders might react to them. For now it's looking like we have price-level S/R to watch for price influence. A retest of this week's lows, near 1867, is probable (until I see something more bullish) and we could see a drop down to the October 2014 intraday low near 1820. So far the March 2014 high and October 2014 weekly closing low near 1885 has been acting as support as price consolidates the sharp decline into Monday morning's low. If the consolidation is followed by another leg down (again, it could be for just another retest of this week's lows) we could have a good setup for a strong bounce to a high in mid-September. I think THE high is now in place and therefore a high bounce in September would make for a very good trade setup on the short side for the next, and stronger, leg down.
S&P 500, SPX, Weekly chart
Looking at the ES (S&P emini futures) and cash index patterns I've been thinking about the potential for a sideways triangle (descending for SPX, ascending for ES) to play out this week and then the retest/new low early next week. If SPX rallies above 1954 that would be a little more bullish, in which case I'd look for Monday's gap to be closed, near 1972, and potentially up to 2008, for a higher a-b-c bounce off Monday's low. Assuming we'll get another leg down by early next week I'll then be looking for a setup to get long for a big bounce into September. Upside projections for the bounce are anywhere from 1980 to 2040 (I'll get a tighter projection as the bounce pattern (assuming we'll get it) develops.
S&P 500, SPX, Daily chart
Key Levels for SPX:
- Cautiously bullish above 2010
- bearish below 1885
The 30-min chart below shows a couple of ideas for a correction pattern off Monday's low. Sorry for the crowded ideas but there are multiple possibilities and I've narrowed it down to "only" three. The descending triangle calls for another down-up sequence into next Monday (maybe finishing this Friday) and then another leg down to complete a 5-wave move down. There are potentially a couple more stair-steps lower for the market over the next couple of weeks, perhaps putting in a tradeable bottom in mid-September instead of a top, but I can only evaluate that once we get more price action. For now the bullish divergences suggest the bears be cautious here. If the sideways triangle drawn on the chart doesn't hold price down, watch for a possible high for the bounce at gap closure, near 1972 and then a projection up to 2008, which is where the c-wave of a larger a-b-c bounce would be 162% of the a-wave (the first leg up from Monday morning's low). Above 2010 would be more bullish but again, potentially only for a higher bounce before turning back down.
S&P 500, SPX, 30-min chart
It's a battle of the trend lines when I look at the DOW's weekly and daily charts. The weekly chart below shows the DOW has created a strongly bullish hammer at support following Monday's low at 15370. Support includes its February 2014 low at 15340, the 38% retracement of its October 2011 - May 2015 rally at 15315 and its 200-week MA at 15290. On Monday it broke its uptrend line from March 2009 - October 2011, arguably THE uptrend line defining the bull market since 2009. The line is currently near 16300 so for all intents and purposes it has now recovered the line with today's rally. Now all it has to do is hold it or climb higher for the important weekly close. A weekly close above 16300 with that bullish hammer (similar but larger than the one at the October 2014 low) would get a lot of traders interested in buying, with the expectation of a new rally leg to new highs. Who's to argue they're wrong?
Dow Industrials, INDU, Weekly chart
The daily chart also shows the bounce off Monday's low but now it looks like it could be a back-test of support-turned-resistance near 16300. A selloff from here would create a bearish kiss goodbye and a strong sell signal. Different from what I'm showing on SPX, for the DOW I'm tracking the idea that the decline into Monday's low was only the completion of one of the nested 3rd waves in the wave count. It says we'll see the DOW stair-step lower into mid-September before making a good tradeable bottom, perhaps down near 14380 for a 50% retracement of its 2011-2015 rally. Note that the decline from here, if it follows something like I've depicted, would be a whippy affair and difficult to trade. Only after hitting the bottom would we have a very good setup for a trade on the long side into November and then a monster decline into the new year.
Dow Industrials, INDU, Daily chart
Key Levels for DOW:
- bullish above 17,100
- bearish below 15,300
NDX has had one of the stronger bounces off Monday's low and if it can close above 4250 (today's high was near 4228) it will close back above its broken uptrend line from November 2012. This is again using the arithmetic price scale because all of its uptrend lines have been broken when viewed with the log scale price, including the one from March 2009 - November 2012. If the bulls can accomplish that with a Friday close I'll have more respect for the upside but for now I'm not so sure the bears are done. A stair-step move lower could see NDX down to the 3600 area by mid-September before it will be ready for a higher bounce.
Nasdaq-100, NDX, Daily chart
Key Levels for NDX:
- bullish above 4385
- stay bearish below 4250
The semiconductor index is always a good one to keep an eye on, not just for the techs but for a sense about the broader economy. With chips in just about all consumer products these days, how the semis are doing is a good reflection of how the economy is doing. While the stock market has been disconnected from reality (the economy) for a long time, eventually it will prove itself to be a good indicator. At the moment, as the weekly chart of the SOX shows, Monday it gapped down and opened at price-level support near 545, breaking its uptrend line from November 2008 - November 2012, near 583 in the process. That's a major bull-market trend line and breaking it is bad news for the bull trend. But as with the others, it will be the weekly closing price that matters and the trend line will be near 585 on Friday. Today's bounce brought it back up to the line (closing slightly above it with the little spurt higher into the close. Each day this week it has bounced back up to the trend line and while today's bounce looks more bullish, it will need follow through on Thursday or Friday. A break, and hold, above 585 would be bullish (leaving a failed breakdown) while a drop back below 545 could result in at least a test of its 200-week MA near 518 next week.
Semiconductor index, SOX, Weekly chart
I'm showing a downside pattern for the RUT that is similar to the DOW's -- both show the market will continue to stair-step lower into mid-September before setting up a much larger bounce correction. The RUT could drop down to its October 2014 low at 1040 or perhaps a little lower.
Russell-2000, RUT, Daily chart
Key Levels for RUT:
- bullish above 1156
- bearish below 1080
Bonds have had a wild ride in the past week as well and it's been a choppy ride since Treasuries bottomed in June. The rally in bond prices has forced yields back down and as can be seen on the TNX (10-year yield) weekly chart below, it broke its uptrend line from February-April (on August 11th), made repeated attempts to get back above the line and then dropped sharply last week, firmly breaking below its 50- and 200-week MAs in the process. This week's rally has now brought it back up to the MAs as well as its downtrend line from December 2013 - September 2014, all crossing near 2.18%. Up near 2.4% is its broken uptrend line from February. A turn back down here would be bearish for yields and bullish for bond prices, which could coincide with another selloff in the stock market. Keep an eye on the bonds. TNX needs to get above 2.4% to at least lend a helping hand to the stock market.
10-year Yield, TNX, Weekly chart
The U.S. dollar spiked down on Monday, with all the turmoil going on in China and other currency issues, and dropped below the top of its parallel up-channel from 2008-2011, which it broke above in January and has been using as support since then. It had also dropped below its 50-week MA at 93.41 but has had a strong recovery the past two days. Based on Monday's low I've modified the expected consolidation pattern from a sideways triangle to more of a shallow descending wedge for a consolidation pattern that should continue through the rest of this year before rallying early next year.
U.S. Dollar contract, DX, Weekly chart
Much of the blame for the diver lower in commodity prices is placed on the rallying U.S. dollar (most commodities are priced in dollars). But as the dollar has traded sideways since March, and trades lower than where it was in March, you can see commodity prices, as reflected in the Bloomberg Commodity index, have continued to sink lower. It's the ol' supply-demand equation and clearly demand has dropped and prices are chasing it lower in hopes of sparking more buying. I don't think we've seen the lows yet but prices are probably near or at a level that will create a multi-month bounce/consolidation before heading lower. Between the bottom of a parallel down-channel for the leg down from April 2014 and the December 2001 low at 85.38 we could see the start of a bigger bounce (in time if not price). I do see a short-term pattern suggesting only a small bounce and then a low near 82 by October before the bigger bounce gets started but I think at this point it could be risky chasing commodity prices lower.
Bloomberg Commodity index, DJUBS, Weekly chart
Gold's rally into Monday's high achieved a projection at 1162.50 (with a high at 1169.80) for two equal legs up from July 24th for an a-b-c bounce correction to its decline. Last week's rally saw gold break above its price-level S/R near 1142, which it had broken below in mid-July, and that was looking bullish. But at Monday's high it tagged its downtrend line from January-May and has done a quick turnaround this week and back below 1142. As I'll show further below with the silver chart, it was not confirming gold's rally and gave traders a heads up that gold's rally might not last. My expectation is that gold will work its way lower toward 1000, if not 900, this year and so far I'm not seeing anything to change my opinion about that. We could see a higher bounce before heading lower, but only if silver participates.
Gold continuous contract, GC, Weekly chart
As mentioned above, while gold was breaking above its shelf of support, near 1142, silver was not doing the same thing. There's a slight uptrend to the line of support from November 2014, near 15.50, that was holding silver down and today it firmly broke longer-term price-level S/R near 14.65, which started back in 2006. If it's not quickly recovered we'll probably see silver drop down to the price projections (based on previous consolidation patterns, as noted on the chart) in the $12 area. A drop down to the bottom of a parallel down-channel from the April 2011 high could see silver trading closer to 11 before the end of the year.
Silver continuous contract, SI, Weekly chart
I've been watching the 127% Fib price extension of oil's previous rally (3-wave bounce from January to May), which is at 38.41. This is a common reversal level if the prior trend is not still in force. I've believed the leg down into the January low would be followed by a sideways consolidation and therefore the 127% extension should hold as support if the downtrend is not still in progress. On Monday it did a minor break below that level but then recovered the next day. If the consolidation pattern is going to hold, perhaps something like what I've depicted on its weekly chart, we should see oil start to rally. But as shown with the light red dashed line, if we get a small bounce and then a minor new low we could then see the start of a larger bounce/rally. As for the downside, there's still the potential for a drop down to its January 2009 low at 33.20. The commodity index has already broken well below its equivalent 2009 level and is testing its December 2001 low. The equivalent low for oil is just above 17, more than a 50% haircut from here (gulp). That would likely not be at all helpful to stock market bulls.
Oil continuous contract, CL, Weekly chart
The market is not paying much attention to economic reports, although there was a slight bump higher after this morning's 8:30 Durable Goods reports. They were better than expected although at +2.0% it was less than half of June's +4.1%, which was revised higher from the originally reported +3.4%. Ex-transportation the number was only marginally better than expected, +0.6% vs. +0.4%. The big number before the bell Thursday morning will be the GDP 2nd estimate, which is expected to improve to 3.1% for Q2 vs. 2.3% in Q1. We'll see...
Economic reports and Summary
The strong decline into Monday's low hasn't been followed by the typical v-bottom reversal we've seen so often in the past. On the weekly chart it looks the same but on the daily charts you can see the difference and it could be a key difference. The technical damage to the charts is significant. The oversold readings (and extreme move in the VIX) could be worked off in a sideways consolidation and then prices head lower again. There is of course the potential for the weekly v-bottom reversal to carry the day for bulls but at the moment there's a caution flag on the track.
Severe spikes to the downside in a bear market (yet to proven we've entered one but I believe we have) are often followed by strong spikes to the upside, usually on news about some kind of action that will save the market. Once it's realized the save isn't going to work (and short covering stops) the bear mauls all those who dared to try to catch falling knives. I don't know if we're there yet but the risk is high for either direction at the moment with all the violent price swings we're seeing, much of which is occurring during the overnight sessions and creating large gap openings for us. Holding a position overnight is dangerous for your health, especially if you find yourself not sleeping well -- lighten up your position and try not to be involved in the market every second of the day. When we see price action like this the best trade is often cash. It's a position, even if you don't like that one.
If prices continue to chop up and down then I would expect another leg lower. I show a stair-step pattern lower for several indexes into mid-September, or we might see only minor new lows before setting up a stronger bounce into mid-September. We should get a better sense of which it will be by the end of the week. But the crash flag is still flying and that means huge risk to the downside is still with us. Stay safe.
Good luck and I'll be back with you next Wednesday.
Keene H. Little, CMT
In the end everything works out and if it doesn't work out, it is not the end. Old Indian Saying