Wednesday, January 6, 2010

Treasury Bonds In The Balance

As I have been chronicling for better than a year now, longer term Treasury yields have a high likelihood of undergoing a secular trend change from down to up. See figure 1 a monthly chart of the yield on the 10 year Treasury (symbol: $TNX.X), which has served as our proxy for the long bond.

Figure 1. $TNX.X/ monthly

The indicator in the lower panel is the "next big thing", and this suggests that yields have those technical characteristics of an asset that are typically seen at the bottom of a prolonged downtrend just prior to a secular trend change. If we go back about 9 months, we have the break of the down sloping trend line; this is at point 1 on the chart. This is bullish especially within the context of the "next big thing" indicator being in a position where trend changes occur. A month later at point 2, prices closed above the prior pivot low point at a yield of 3.432%. This also is bullish. Prices did not hold beyond a month, and yields fell back below our pivot point of 3.432%. This inflection point was resistance for 5 months until yields closed the year back above the 3.432% mark and above the down sloping orange trend line; this is point 3. Once again, this is bullish price action.

However, turning to the weekly chart, I must give some pause. See figure 2. The indicator in the lower panel looks at the 52 week rate of change of the yield on the 10 year Treasury bond, and that value is then wrapped in dynamic trading bands. These trading bands use a 104 week look back period, and readings outside of the bands are 2 standard deviations outside the norm. The current indicator value is not only outside the bands, but it is the highest value in almost 50 years! That is extreme.

Figure 2. $TNX.X/ weekly

Now here is the trick and this is key: if you have been around the markets for awhile, extremes can mean revert or they can indicate a break out and the start of a new trend. So at this juncture all I have is an issue that has moved very far and very fast over the past 52 weeks. I am very mindful of the bigger picture on the monthly chart and the significance of trend line breaks when the "next big thing" is in its current position. Furthermore, let me remind you that not all price moves end up being tradeable. Refer back to figure 1 where I labeled the 2004 to 2007 up trend in Treasury yields; despite the persistence to go higher, this was a choppy trend. So my first concern is the extreme overbought reading, but as stated, overbought can be good too.

My second concern is seen in figure 3 another weekly chart of the 10 year yield. The pink labeled price bars represent negative divergence bars between price and an oscillator that measures price. The indicator in the lower panel measures the number of negative divergence bars occurring over a certain time period. When the indicator is red it is identifying a cluster of negative divergence bars, which typically indicates slowing upside momentum and the possibility of a market top. In fact, this was pretty much the case for yields on the 10 year Treasury from the 1980's to the current time, and as you can see, the last occurrence of this cluster was in August, 2008 prior to yields unraveling.

Figure 3. $TNX.X/ weekly

Three things are important about this cluster of negative divergences: 1) the current yield is right at the highs of this cluster of negative divergences; yields haven't rolled over but they have not broken above this resistance area either; 2) in the 1970's, which was the last bear market in bonds (or when yields went appreciably higher) these clusters did not end the trend but only slowed it down; in a previous article, I remarked that we would know that the trend for higher Treasury yields is for real if this resistance zone was easily taken out; 3) breaks over resistance areas formed by negative divergences can lead to accelerated price moves; we have seen this over and over again in the equity markets for the better part of the past 6 months.

My third concern is bond market sentiment, and this can be seen in figure 4 a weekly chart of the yield on the 10 year Treasury bond. The indicator in the lower panel looks for extremes in the Market Vane Bullish Consensus data for Treasury Bonds. When the value is low and below the lower trading band, it generally signifies a top in Treasury yields. That is how it is suppose to work unless it is 1994 to 1995 or 1999 to 2000 when yields rose in spite of the very bearish sentiment for bonds. The current value is below the lower trading band, and this means investors are too bearish on bonds or too bullish on yields.

Figure 4. $TNX.X/ weekly

But as stated above and as we know from this year's stock market, extremes in sentiment don't always lead to turning points, and in fact, the increase in yields during the 1994 to 1995 and from 1999 to 2000 saw lots of bullishness for yields. Yet, yields still went higher!

So let's stop and summarize the technical picture. Yields on the 10 year Treasury have moved an extreme degree and are currently hitting a "key" resistance area formed by multiple negative divergences. Sentiment, which works "most" of the time, is too bullish on yields. None of this suggests an imminent reversal, but it is the current state of affairs. The longer term monthly picture is more bullish, but in light of the weekly picture, I wonder if we are going to get a picture similar to 2004 - 2007 that was characterized by spiking yields followed by months of downward yield pressure.

On a fundamental basis, higher yields appear to be implying an improving economy although most would agree that higher yields are bound to choke off any recovery. Higher yields could also remain persistent as long as Dollar devaluation and inflation remain a concern, but the downtrend in the Dollar appears to have played itself out for now, and in an economy still under the weight of a credit contraction and deleveraging, most would agree that inflation is far off in the future. So may be higher yields are signaling an improving economy.

So the question in my mind is this: how do we resolve these issues?

For this I need to introduce what I call the "key" pivot point. Key pivot points are special pivot points as they are a pivot point low occurring at a time when investor sentiment is bearish (i.e, bull signal). Why are these key areas? Well think about it for a second. Sentiment gets extremes and prices should bounce higher and a pivot low point is formed on the chart. Typically, such price action will represent the low. However, failures do occur despite the extremes in sentiment, and prices can trade back below these key pivot points. It is these failures at the lows that lead to a reversal in the trend catching traders and investors off guard.

Let's look at a weekly chart (figure 5) of the i-Shares Lehman 20+ Year Treasury Bond Fund (symbol: TLT). The indicator in the lower panel looks at the Market Vane Bullish Consensus data for bonds, and the current reading is below the lower trading band suggesting bearishness towards longer term Treasury bonds. That is, investors are betting on higher yields. The black dots are pivot low points and the black dots within the yellow dots are our special, key pivot points.

Figure 5. TLT/ weekly

The key pivot point at point 1 was the bottom for bonds back in 2003, and this was retested at point 2 in 2004. In fact, prices at point 2 closed below the key pivot point from point 1 (prior support) suggesting lower yields, but this level was quickly re-captured and bonds moved much higher over the next year. Point 3 is a pair of key pivot points that are higher than key pivot point 2 suggesting a higher low and an intact bull market for bonds. Point 4 was a retest leading to the final blow off in TLT in early 2009.

Now we come to the 3 key pivots on the right side of the price chart. The close below the key pivot point at point 5 is bearish, but history would have us believe that caution is in order. The close below the key pivot point at point 6 is even more bearish for bonds, and the same can be said for the close below the key pivot point at point 7. In fact, a close below 3 key pivot points is a good sign of a secular trend change! (See this article on the Dollar Index.)

The picture is not looking good for bonds. Expect TLT to continue on a downward trend.

What would get me to change my tune on TLT? A weekly close above the key pivot point at point 7 at $89.91. Expect prices to flirt with this level. A weekly close above the key pivot point at point 6 ($92.15) would be a reason to become very bullish on TLT again.

Lastly, let's look at the i-Share Lehman 7-10 Year Treasury Bond Fund (symbol: IEF). See figure 6. The key pivot point at point 2 traded below the point 1 key pivot point, but once this level was recaptured, IEF moved higher. The same could be said for the point 3 key pivot point. The key pivot point at point 4 did not close below point 3, and this proved to be bullish. Once again, focus on the right side of the chart and we note closes below the prior two key pivot points, and this is bearish. Price has yet to close below the bottom most key pivot point at 88.62. A close below this level suggests a bearish picture for bonds (or higher yields) and IEF. If this level holds (or is recaptured later following a break below), then I would become very bullish on IEF on a close above the middle pivot point at $89.79.

Figure 6. IEF/ weekly


How am I playing it? I am still long IEF with a cost basis of $89.85; the position represents about 44% of my portfolio. I have recently added to this position at $88.62 recognizing that I will be gone from this trade on a weekly close below $88.62. It would have been nice to get out 5 weeks ago when prices closed below the most recent key pivot point, but things didn't work out that way. I am prepared to exit the trade on a close below $88.62. If prices close back above this level in the following week, I will likely re-initiate my long position. It may be a pain moving in and out of a position in this manner, but I look at it as good insurance.

From this perspective, Treasury bonds are at risk for a significant down trend. To put it another way, Treasury yields could finally be going significantly higher. The price action around these key pivot points needs to be followed closely as we are in an area where new trends develop.

6 comments:

Johnny G. said...

Excellent, thorough analysis. Friday's NonFarm Payrolls number may resolve the issue in the near term. Due to seasonal adjustments, we could get a very high number (+100,000 or so), which could send yields much higher.

D-man said...

Top notch!

I want to make a comment on chart 3 (weekly yields). And I'll make this with the chart of LT S&P 500 you analyzed the other day.

The idea is you get a top in the market after a couple of months of market movements drawing 2 or more clusters of negative divergences; this is the typical M top.

Although chart 3 is weekly and not monthly, we can formulate a conclusion for the intermediate term.

We can clearly see we have a cluster of negative divergences for the first part of the M formation (if this will prove to be an intermediate M top); now if the market draws another cluster of negative divergences for the second part of the M (or double top, whatever you wish) and rolls over, yields will go lower in 2010.

The pivot of the M is at 31 as drawn on the chart. Subtracting this from the top at 38 we get a target for TNX

31 - (38 - 31) = 24

which means we will revisit the low yields seen back in 2008.

LT it's another story as you say.

What this means for the economy? Another round of "deflation trade"; of course this is very difficult to predict as there is so much intervention out there by the FED, Congress (I assume that's why analyzing these charts is so difficult and the charts are so ugly). It's basically a battle between the true deflationary forces in the economy and the printing presses of central banks; imo, the markets are rigged, but that's of course personal opinion and I don't think we'll find that out anyway :)

All in all, these are very ugly and choppy charts.

thanks again for this detailed analysis!

Guy M. Lerner said...

The charts are ugly; but with a longer term view, I try to stay out of the noise.

This is important to me and I spend a lot of time on bonds/ yields for several reasons: 1) non correlated diversified asset; my biggest fear is that everything else runs together including commodities, stocks, reits, emerging markets; 2) if I can the secular change part correct this could be a big trade

By the way, I think the Dollar goes higher!!! More later

D-man said...

btw Guy, everyone seems to put on the table the "yield curve" (or the spread between short and long term yields) and predict with it the economy. The curve, based on the last 50 years, shows big recovery; never in the last 50 years there was recession or weak activity with such a curve as today. The only problem is these times are quite extraordinary and the FED tries hard to manipulate the curve; I know "this time is different" never works and maybe there will indeed be strong economic activity.

So while an economic recovery might drive stocks and yields higher, the problem with stocks is they are quite expensive (sub 2% yields), even if companies start to repay some dividends.

See the link
http://www.hussman.net/wmc/wmc091221.htm

Guy M. Lerner said...

My only caution regarding the yield curve is this: most fundamental data tends to fail at some point in time.

Let me give you an example. I remember reading an investing book in 1999 suggesting that all you had to do to make money in the stock market was follow the Fed. In general, good advice. Ahh, but then came 2000 and if you bet on the market because the Fed lowered rates you would have been hammered in 2000, 2001, 2002. What used to work no longer worked. If you followed the old line of "don't fight the fed" you lost a lot of money

RB said...

The breakdown in VFITX seems even more pronounced than in IEF.