Tag Archives: volatility

Narrow ranges and light volumes are the subject of a recent blog post by Brett Steenbarger in which he talks about the impact of such market conditions on trader performance. Brett’s specific focus is on stocks, but as I’ve written before, low volatility has been a feature of multiple markets such as interest rates and the USD.

Specific attention has recently turned to the VIX. As can be seen in the daily chart below, the so-called Fear Index has reached its lowest level since the first part of 2013. When the VIX is low it tends to make some market observers nervous, getting them looking for stocks to take a tumble.

VIX and S&P 500

We’ve seen quite a few low VIX readings in the last couple of years, and they have indeed generally been followed by some kind of market reversal. While those downside moves have at times been quite sharp, they’ve never done any real technical damage and the overall trend has remained positive. In fact, the high volatility episodes have tended to be relatively brief, which is often a good indication of a strong trend. Still, a quick 5% drop like we saw the S&P 500 experience early this year is something which can rattle traders- perhaps even more so in a situation like we’re in now where at least part of the market is quite sensitive to the idea that we’re due for a reversal.

While I certainly can envision the S&P 500 experiencing a sharp sell-off at some point in the not too distant future, I’m not overly worried at this point about a major top developing. The market psychology doesn’t really feel excessively positive to me here. I will, however, be keeping an eye on how the ranges in the currency and yield markets eventually resolve, as they will certainly play a part of the future direction – and amplitude – of the stock market’s path forward.

Some noteworthy downside volatility in US stocks in the last couple of weeks has got the global markets quite nervous. Have we seen the top in the market? That risk is certainly there.

The first thing on the chart which jumps out negatively to me is the shooting star candlestick pattern on the weekly chart from last week. That was the result of the market making a new high, then failing and actually finishing low on the period. That’s visible in the weekly S&P 500 chart below.

Weekly S&P Chart - click to enlarge

The other thing I’m looking at which isn’t yet decisively bullish or bearish yet, but which represents potential either way, is the narrowness of the Bollinger Bands. As the lower plot on the graph above shows, that width is at its narrowest for quite some time – lower even than it got in late 2012 before the S&P 500 went on a 100 point rally over the next few months.

Narrow Bands worth both ways, of course. As such, if the market were to fall below about 1775 and get the Bands widening as a result, the implications would be negative. We are already seeing in the daily chart time frame that a relatively narrow Band situation there, combined with the break below the bottom of a key range, is seeing the S&P flip into a negative trend mode.

Daily S&P Chart - click to enlarge

It looks very likely from the daily chart set up that we’ll see the S&P 500 test the lower weekly time frame Bollinger Band. That would be a meaningful move at this point, so on a first go we may not quite see 1775 broken at this stage. That, however, could just set things up for a move through that level, and the widening of the Bollinger Bands which would happen as a result, signalling a likely longer-term bear trend move. Even then, though, the 1650-1700 area where the market ranged before would be a sticking point.

A while back, motivated by persistent misinformation being presented in the media, I did a quick study of volatility in various markets. It was something I repeated here last year using weekly rather than daily data. Markets have had lots of different things happening recently, so I decided to re-run the daily data study, this time using 5 years of data rather than just the 1 year I went with the first time around. Also, rather than showing the figures in tabular format, I’ve decided to make things more visual and put the results in a pair of charts.

Here is the first one which compares four exchange rates, four major US stock indices, eight big cap common stocks, four major commodities, and four key US interest rate instruments (using futures for the latter two groups). It looks at volatility from the perspective of the standard deviation of daily % returns. This basically gives us an idea of how much of a change we see in each market on a given day.

The results are consistent which my prior analysis. The interest rate and exchange rate markets are noticeably lower in volatility than commodities, stocks, and stock indices. The major forex pairs move roughly about the same amount as longer maturity fixed income instruments like T-Bonds (using a price basis rather than an interest rate basis).

The second chart uses average daily ranges as the comparison point. The ranges are calculated as (High – Low)/Prior Day’s Close. This allows us to look at percentages for all markets so we’re comparing apples to apples.

There’s a bit of shuffling around in the order in which the markets rank when looking at ranges rather than returns, but generally the pattern is the same. Interest rate and exchange rate markets fall on the low end of the scale while individual stocks and commodities are on the high end.

The first reason for showing these images to you is so the next time someone tells you how risky the forex market is you can show them the comparison and ask them if they want to reconsider.

The other reason I bring this subject up is to provide a better understanding of volatility across markets , which factors into things like bid/ask spreads, margin requirements, and the like. This should help in your investment asset allocation process – or at least to have some insight into how different markets move if you’re just focusing on only one or two. Of course once you start applying leverage you can create a situation where any market can result in a very volatile account equity line.

Just about two months ago I wrote about my expectations for an expansion in volatility in the dollar. That was based on looking at the weekly USD Index chart and observing very narrow Bollinger Bands, which is often a precursor to the start of a new trend in the markets. As we can see, the greenback has indeed broken clear of the range it started the year in, and of late has reached within striking distance of the 2012 peak.

We can see from the Band Width Indicator in the sub-plot above that the Bands are now in about the middle range of how wide they have been over the last couple of years. That means there is yet some room for further extension to the trend (thought it should be noted that sustained trends often seen the Bands start to narrow).

The daily chart gives us an indication that we just might see that fairly soon. Notice how in that time frame the Bands have reached their lowest levels in the last 12 months. That, in and of itself, is not a directional indication, however. It merely tells us that we probably don’t have long before the recent consolidation in the USD gives way.

So what can we glean about potential direction from the dollar pairs?

If we look at EUR/USD we can see the market has already extended its trend lower and turned a low Bollinger Band situation up.

The USD/JPY situation is quite different. Here we have a market which has already retraced quite a bit from its highs and has been developing a topping type of pattern overall of late.

Similarly, GBP/USD has also been working back higher, showing indications of a trend change.

AUD/USD has been moving higher for some time now. It does face important overhead resistance from the highs near the start of the year, however.

Meanwhile, USD/CAD shows a pattern not dissimilar to that of USD/JPY and GBP/USD in that the greenback has been weakening noticeably of late.

Taken together, we have a picture telling us that the only reason the USD Index is holding near its higher levels is the weakness in the euro. As a result, if the single currency develops some strength – even if it just takes a pause to consolidate the recent downtrend – we could see the index drop. That is not at all contrary to what we see on the USD daily and weekly charts. It would be quite easy for the greenback to retrace back toward the Q4 highs broken earlier this year. That would be important support now.

You may recall my post from a few weeks ago talking about how the strong stock markets may be signalling a turn in the dollar. In a way, we’ve already seen it happen as equity indices have continued to move higher. Now it’s just a question of whether the euro can stabilize and put the USD Index under pressure as a result.

If you are one of those forex traders who has been moaning about the lack of volatility in exchange rates for the last while now, you may not have to wait long for better days. Conversely, if you have been quite happy and successful in trading the relatively quiet markets of the last few months (likely based on range-oriented strategies), now may be the time to start preparing for more challenging days to come. I think we’re just about due for a shift in the markets.

Why do I say that? Give this weekly chart of the USD Index a look.

Take particular note of the bottom sub-chart, which shows the relative width of the Bollinger Bands. That line is at its lowest level for a number of years. The implication is that we are not far away from some kind of significant range expansion in the Index.

Interestingly, we have already seen the start of that increased volatility in the way the euro and yen have both made sharp moves. USD/JPY has been up strongly the last few months, but EUR/USD has only just started moving higher (rather cancelling each other out in terms of the USD).

If we look to the other majors, however, there is still a strong range bias. Here are the charts for AUD/USD and USD/CAD.

GBP/USD and USD/CHF are both a bit more frisky, though.

It may be a bit much to expect a lot more upside volatility in USD/JPY, but there’s nothing to say we couldn’t see that rate have a counter-trend period which is just as violent. The question for a general USD volatility expansion is whether it comes from weakness or strength. The index has been kept in check because other currencies have been the focus recently thanks to Japanese monetary policy, a weak UK economy, and on-going Euro Zone issues. The first chart above, though, suggests we should be looking for something to develop which is USD-specific in the weeks ahead. Chances are, we’ll see that linked with comparable moves in the US stock and bond markets, and perhaps commodities as well.

One of the subject that’s come up in market discussion of late is the low level of the VIX, the so-called “fear index”. It has reached its lowest levels since back in the early days of the Financial Crisis in 2007. The chart below shows how the VIX has moved up and down since 1992.

The question which comes to my mind is whether we’re seeing a pattern similar to the one following the 1998-2003 period of elevated VIX value where the index retraces back to a lower level for a while as it did in the middle 1990s. The action in the markets in the early 2000s and the likes of Enron knocked a lot of individual investors out of stocks, just as recent developments have done. That makes a case for a similar kind of shift in volatility.

Just for the sake of comparison, I think it’s worth looking at volatility in other ways as well to see how things are playing out. The chart below shows the S&P 500 over the same time frame as the VIX chart. The two subplots show the relative width of the monthly Bollinger Bands (BWI) and a normalized 14-month reading of Average True Range (N-ATR). These give us a reading on how much movement there is in monthly closing prices and how wide the monthly ranges are respectively.

It is interesting to note that the Bollingers are back to being nearly as narrow as they were in the middle 2000s after working back from getting very wide back in 2009. As with the VIX, there is still some room to work lower to match prior lows, but we’re back into roughly the same range.

The N-ADR reading is a different story, though. In this case we’re nowhere near back to the lows of the middle 2000s and middle 1990s. In fact, N-ADR has been rising the last year! This tells us that while price changes from month to month may be getting smaller, the inter-month volatility remains elevated. Could this be a tip-off?

As a technical analyst I have a major concern with the way the S&P 500 made a lower low on the monthly chart back in 2009. That’s a big negative. Add to that the fact that momentum in the rally since then has backed off, as indicated by the lessening beats of prior highs for recent new highs, and you get reason for concern. If the market cannot overcome the 2007 highs on this rally, we could be in for quite a bit of a tumble. And when stocks tumble, volatility tends to rise quite dramatically. Generally speaking, before the top is put in the N-ATR reading is already on the rise. We’re seeing that now, so it’s definitely worth keeping that in mind, though given that we’re looking at monthly charts here, it may be a while yet before any sort of roll-over takes place.

As a change of pace, I thought I’d use this post to do a little bit of market analysis looking at the market from a different perspective than the ones most often used. This type of analysis focuses more on time spent (or volume transacted) at certain levels rather than looking at simple progression of where it’s been over time as we generally see in bar and candlestick charts.

The chart below shows how EUR/USD has traded since February. Each of the clusters you see represents the distribution of trading over one month’s time. I won’t go too far into the details, but suffice it to say that the fatter a month’s distribution at a given price level, the more days the market traded at that prices, and the thinner the distribution the fewer days the market traded at that level. Think of it this way. If the market spends a lot of time at a price level it indicates agreed upon value in which both buyers and sellers are willing to transaction. Where the market doesn’t not spend much time it indicates rejection by one side or the other – value not agreed upon.

What we can see above is a trio of short, fat distributions for February, March, and April that indicate pretty narrow range trading. Then, in May, we have a long, thin distribution indicating a trend move lower. June was again mainly a consolidative month, but July started off with a trending action, then transitioned into more of a ranging set-up.

The July distribution indicates that things changed in EUR/USD near the beginning of the month and previously accepted value between about 1.2400 and 1.2700 suddenly became rejected. The market then move down to where valued was agreed upon below 1.2400.

Let’s put this in some common parlance. Think of the thin distribution of prices between 1.2350 and 1.2500 or so as a key resistance zone for EUR/USD. Selling interest far exceeded buying interest the last time the market moved through that zone. If the market can work back up there and hold the move it would tells us things have shifted and that buyers are starting to be more interested.

The concern I have, though, is that we don’t have as clear a rejection area to the downside to indicate a price level the sellers clearly found too low and/or where the buyers became much more aggressive. We have to go back to June 2010 to find the last time the market was down this low. Back then there was a final rejection near 1.1900. I think the risk, therefore, is that EUR/USD makes another move down to test those prior rejection lows.

The struggle, though, will be breaking away from the 1.2300 area. As the chart above shows, the market spent a lot of time around there in May/June of 2010. That makes it a significant attraction zone, which we’ve been seeing play out this month. If the market can start to develop more value below 1.2200, though, the odds for a run at 1.19 will increase.

There’s a bit more nuance to this type of market analysis, of course. If you find it interesting, you can learn more about it here.

The forex market continues to make some folks nervous. While there are certainly reasons to be cautious when playing exchange rates, a considerable amount of the nervousness of the average person on the street comes from misinformation. Most notably, they all too often think of the forex market as being highly volatile. I addressed this issue before in Looking at Volatility Across Markets, but I think it’s worth revisiting.

I’ve collected 5 years of weekly values for a number of markets to look at just how volatile they are. Let me first look at the US Dollar Index. Between July 2007 and July 2012 the average weekly range for the USD Index was just 2%. I derived that by taking the distance between each week’s high and low and dividing it by the midpoint for that week [( High – Low)/( (High+Low)/2) ]. At the same time, the standard deviation of weekly closing prices (which gives us an idea of how choppy the market is) was only 4.9% (relative to the average close for the study period).

As you will see, that’s not a lot of volatility.

Let’s start by comparing the USD Index values to those from the major US stock indices.

DJIA: 4.0% average weekly range, 15.0% standard deviation
S&P 500: 4.4% average weekly range, 16.0% standard deviation
NASDAQ 100: 4.7% average weekly range, 20.2% standard deviation
Russell 2000: 5.6% average weekly range, 17.3% standard deviation

As you can see, the major stock indices show considerably more volatility than does the USD Index.

How about individual stocks?

JPM: 9.4% average weekly range, 15.0% standard deviation
IBM: 4.8% average weekly range, 24.6% standard deviation
GE: 7.4% average weekly range, 39.9% standard deviation
XOM: 5.1% average weekly range, 12.0% standard deviation
KO: 4.0% average weekly range, 15.3% standard deviation
AAPL: 7.2% average weekly range, 53.5% standard deviation
KO: 6.2% average weekly range, 18.0% standard deviation

No real surprise to see that individual stocks are pretty volatile by comparison.

Looking at commodities:

CRB Index: 3.9% average weekly range, 18.2% standard deviation
Gold: 4.6% average weekly range, 28.1% standard deviation
Crude Oil: 8.4% average weekly range, 24.2% standard deviation

Here again we see markets with a great deal more volatility than the USD Index. The one place where there is something of a contest is the bond market. The long-date Treasury note/bond ETF is TLT. Looking at its weekly figures I come up with a 3.2% average range and 10.6% standard deviation. That’s considerably less volatility than the other markets and securities shown above, but still not at low as what we have seen the last five years in the USD Index. We would likely have to move down to short-term Treasury securities (like 2yr Notes and T-Bills) to find lower values.

The point of all this is that anyone avoiding the currency market because of the perception that it’s super volatile is operating on a false belief. The figures just don’t back that up.


1 Comment

High frequency trading (HFT) has gotten a lot of press over the last several years. The so-called “Flash Crash” of May 2010 highlighted the impact HFT can have on the markets, though the focus has mainly been on the stock market. No real surprise there. The average person on the street still thinks of stocks first when the subject of trading comes up. That’s not the only place HFT’s influence is being seen, though, as a recent newswire piece indicates.

Here’s the first part of the Dow Jones story (which can be read on the Wall Street Journal site here):

Electronic inter-dealer currencies-trading platform EBS plans to scrap the fifth decimal place on its currency quotes and introduce so-called half-pip pricing ahead of major changes to the system, people familiar with the matter told Dow Jones Newswires Tuesday.

EBS, owned by ICAP PLC (IAP.LN), has been considering a range of options that will change the way investors are allowed to trade on the system in a bid to repair relations with its core banking customer base. EBS shares a dominant position in currency markets with Thomson Reuters (TRI), but it has come under fire from its core bank clients for allowing trading behavior that seemingly favors so-called high-frequency traders in recent years. Now it is seeking to redress that balance.

This article caught my attention because I recently reviewed a book titled Broken Markets on the subject of market structural changes and how HFTs have been able to exploit them. One of those changes is the move to decimalization made by the stock exchanges in the US in 2000. That helped narrow bid/ask spreads, which lowered trading costs. According to the book’s authors, though, it also created a lot more price points for trades to take place, leading to thinner liquidity at any given point, which is something noted in the Dow Jones story from the forex perspective. It also created greater opportunity for HFTs to come in and do their thing (some of which is highly predatory). The introduction of pipettes (fractions of pips) in forex has served the same purpose.

The move by EBS, as motivated by their customers (mainly the major banks – see The Dominant Players in Forex), is to pull that back a bit. The banks are feeling the pressure in their dealing margins, which have already been squeezed considerably. Back when I started in this business the bid/ask spread on USD/JPY was consistently 10 pips and up. It’s more like 2-3 pips these days most of the time, which means the banks who are acting as market makers are making much less profit per trade.  This spread compression, combined with rapid technological development, has been a big factor in the shrinking of the global foreign exchange business. You can understand why the banks wouldn’t want to see those spreads narrowed further, especially if HFTs are grabbing a rising share of the volume.

If EBS makes the move to limit pipettes to only half pips, as proposed, that could have an impact on retail forex broker pricing. Exchange rates cascade down from the inter-bank market to the retail one, so any development at the top end where EBS operates is likely to filter its way to broker platforms. Not that it’s likely to have a major impact on most individual traders’ strategies.

The other thing EBS is looking to cut down on is the kind of quote stuffing that can lead to illusory liquidity. This quote stuffing happens when an HFT submits quotes/orders to the market really only to identify liquidity and feel out price direction. A high percentage of these orders are quickly cancelled. EBS wants to crack down on this, which could have an impact on price action. That may end up being the bigger thing to keep an eye one moving forward. If it tends to smooth out prices we could find forex becoming that much more interesting for that money flowing out of the stock market.

In the end Wednesday, the markets got just about what was expected from the FOMC and Fed boss Ben Bernanke. While a certain notable French bank who shall remain nameless (OK, it was Société Générale) came out with 70% odds of a $600bln round of new quantitative easing (aka QE3), that was an outlier view. Most folks in the fixed income and forex markets (we don’t pay much attention to the stock guys :) ) were looking for a continuation of the Operation Twist program in which the Fed sells short-term treasuries it owns and buys longer-term ones.

These expectations are why in the end the various global markets basically just continued on the course they had already begun earlier in the day, albeit with a little volatility after the FOMC statement and into Bernanke’s press conference. Following the extension of Twist, the Fed chief’s comments about standing ready to do whatever may become necessary were predictable. He’s been saying that for some time now. Why not? It’s true. It’s always true. The Fed will do what it thinks it needs to do when it thinks it needs to do it. Folks seem to read QE3 expectations into that every time he says it, though.

To that end, it occurred to me yesterday that the folks who keep calling for QE at the next FOMC meeting are kind of like the folks who set dates for the end of the world, then when it doesn’t happen they revise to a future date.

The thing that had me sure there was no QE3 coming this week was a comment Bernanke made a little while ago that he was seeing no signs of deflationary risks at present. Deflation risk was a big factor in the justification for QE in prior rounds, so if he’s not seeing that risk now, the odds of QE3 drop despite economic developments. Now, the Fed forecasts released yesterday did feature lower inflation expectations, but nothing leaning toward deflation. That will be something to watch morning forward.

At this stage, the bigger issue at hand is going to be the value of the signals coming from the Treasury market. As I wrote a couple weeks ago, the Fed already owns a large portion of outstanding long-term Treasury paper. The extension of Twist is only going to make that share grow. The bond market guys I work with say basically the Fed will be buying all of the long-dated paper the Treasury issues for the rest of the year. This is going to further shrink the “float” of long-dated securities, which could make the likes of US 10yr yields even more volatile because it will take increasingly smaller volume to move them around.

Considering how correlated USD/JPY tends to be to those rates, the higher volatility in yields could make for some interesting action in that exchange rate. Notice in the chart below how much time the correlation between the two markets is positive and how even when it turns negative it is just briefly and only marginally so. If the 10yr yield becomes less valuable as an indicator due to the Fed’s dominant holdings, we could see the relationship between it and USD/JPY breakdown.

Operation Twist Chart

Also, things could get interesting on the short end of the yield curve as well.

The Fed normally holds a lot of short-dated Treasury paper which it uses in open market operations to keep short-term yields in line with policy. The Twist operation has already seen a lot of that paper sold as the Fed has bought long-dated securities. The expectation in the bond market is that the Twist extension will result in the Fed not having any shorter-term paper left. That could create some interesting dynamics at the front end of the yield curve. Considering how important overnight interest rates are to currency exchanges rates, there is the prospect of some periods of unusual activity in the months ahead. As a result, it will be worth keeping track of what the Fed is doing.

This is one of those times when understanding structural elements of the markets can be important.