Trader’s Narrative

The mad rush into US government treasury bonds has pushed their yields to never before imagined levels. But according to the simple 1 year rate of change, we may be close or have already seen the end of this “bubble”, as this long term chart shows:

Although I think the bond market will return to its senses soon enough, US government bonds are not in a real “bubble”. But the extra-ordinary demand for US treasury bonds is at least partially responsible for the strength of the US dollar. Which I’m sure itself is surprising and confounding the gold bugs more than anyone else. After all, how can a near meltdown of the world’s financial system result in gold falling and the dollar going up? Isn’t that the exact opposite of what a sane person would expect to happen?

This recent chapter in financial history is chock full of unprecedented extremes and “Black Swans“. Among them, the yield inversion between equities and treasury bonds (Bloomberg):

Something that we haven’t seen in 50 years. And something that was just as jarring when it was witnessed in 1958. The sharp drop in rates is half the explanation, the other half is the dramatic rise in stock dividend yields.

Most market watchers are perplexed about what this means, if anything. The only explanation offered which makes some sense comes from Cliff Asness of AQR Capital:

From the 19th century through the mid-20th century, the dividend yield (dividends/price) and earnings yield (earnings/price) on stocks generally exceeded the yield on long-term U.S. government bonds, usually by a substantial margin. Since the mid-20th century, however, the situation has radically changed. In addressing this situation, I argue that the difference between stock yields and bond yields is driven by the long-run difference in volatility between stocks and bonds. This model fits 1871-1998 data extremely well. Moreover, it explains the currently low stock market dividend and earnings yields. Many authors have found that although both stock yields forecast stock returns, they generally have more forecasting power for long horizons. I found, using data up to May 1998, that the portion of dividend and earnings yields explained by the model presented here has predictive power only over the long term whereas the portion not explained by the model has power largely over the short term.

More about the relationship between dividend yields and US treasury bonds:

John Mauldin: Two Little-Noted Features Of The Markets And The Economy
Bloomberg: S&P 500 Payout Tops Bond Yield, a First Since ‘58: Chart of Day
Barron’s: Reversal of Fortunes Between Stocks & Bonds
Mark Hulbert: Stocks vs. bonds

Well, do ya? If not, let me refresh your memory.

Last year, on June 4th, Morgan Stanley’s market strategists released a message telling their clients of a “Full House” sell signal. The note didn’t attract any attention until June 6th when it ignited a sell-off in the European bourses and jumped the pond to North American exchanges.

The three alarms were fundamentals, valuation and risk indicators. The latter two had already been lit and with the higher bond yields and new stats on manufacturing orders, the fundamental indicator completed the triple threat.

First of all, the note was from a respected Wall Street firm and second, it had a fierce record: it had only happened 5 other times since 1980, with each instance garnering an average loss of 15% in the following six months.

There was some confusion because the Morgan Stanley US strategist had a different take and it wasn’t clear if the call was for just Europe or whether Morgan Stanley thought it had implications for other markets (it wasn’t and it hadn’t). But before things could be cleared up, the media had latched on to the headline and the damage was done.

So why am I bringing this up after so much time has passed? It is only after time that we can look at it with some perspective and perhaps, learn something for the future.

It is so easy to get caught up in the current market-tell, whether it is a Fed decision, ISM report or whatever. What isn’t so easy is to keep one’s head and remember that these effluvia are part and parcel of the stock market’s fog of war. Also, experts are hazardous to your financial health.

Morgan Stanley’s note did spook the market into selling off significantly. But the general sentiment was too bearish already for any real damage. We had a lukewarm retest in July which didn’t reach the June lows and away we went.

It is a cliché but that doesn’t mean it isn’t true: the stock market climbs a wall of worry. Looking at the chart above it’s pretty obvious that the 3 alarm sell signal was just a brick in the wall.

See any bricks today?

UPDATE: If you’ve read the above with care, you’ve noticed that I made a mistake in the years (2006 instead of 2007). The 3 Alarm Sell Signal was in fact a very good one and this post was a silly but honest mistake.

According to the simple stock market timing method which relies on the 30 day rate of change of the 10 year treasury bonds, we are very oversold and should expect a rebound here.

This system is somewhat better at finding tops than bottoms, but it is a good general indicator to throw in the pot:

The rate of change (bottom panel) is really off the charts. The chart doesn’t go back that far but we haven’t had a reading this low since May 2003 when the S&P 500 was at 1000; lifting off a triple bottom and escaping from the clutches of the bear market.

It wasn’t that long ago when people were running around pulling their hair out because of a swift sell off in the bond market. As yields spiked in the 10 year and 30 year notes, a lot of attention was invested in trying to explain what this meant.

In early July I showed one simple indicator based on the 10 year Treasury Notes and how it has a very good record of finding market tops - something most indicators are loath to do.

Here is the indicator again, going back three years:

As you can see, this is able to not only find tops, but bottoms as well. Whenever the rate of change of the 10 year T-Notes drops into negative territory, we start to see a high probability of the equity market bottoming. It isn’t perfect, as you can see. It missed the October 2005 bottom (or was early depending on how you want to think about it).

The rational argument is that as bond yields fall, equities become more appealing compared to bonds. And so funds flow from bonds to stocks. Although not simple, it is both logical and fairly consistent. What else can you ask from an indicator?

With the rising yields (falling prices) in the long end of the bond market, the yield curve is back to normal. Short term yields (90 day) are at around 4.80% and long term yields (30 year) pushing 5.25%.

The 10 year bond, what everyone seems to be watching these days, gapped up and closed very strong. This sort of move, although seemingly strong, is usually indicative of the last sprint before exhaustion. Which is why such gaps are called “exhaustion gaps” - but only in hindsight since its 20/20

So while the 10 year bond yields are rising, perhaps saying that the Fed won’t cut… the very short term bond market is staying stubbornly below the Fed funds rate. Only one of them will be proven right.

In any case, all this attention on the bond market got me thinking about simple mechanical trading systems that I mentioned before. I wondered if the bond market could be used to time the stock market in the intermediate term.

Tops are extremely difficult to identify. I haven’t really found many reliable indicators for the job. Put/call ratios and the myriad variations of that sort of data? Helpful. Sentiment surveys? At times. But what about bond yields?

I looked at the 10 year bond yields and did a very simple 30 day rate of change (ROC) calculation and compared it to the S&P 500 index. Surprisingly, it did a very good job of pointing out market tops. Take a look.

Click to Enlarge Graph

As you can see, when the ROC was approaching 9% things got a bit queasy for the bulls. The most recent case being just a few days ago.

The one glaring mistake it made was in October 2005 where it flagged the exact intermediate bottom as a top. But we can easily avoid that sort of thing by having a simple condition that the signal is only good if the market has been rising in the most immediate past.

And eventhough I wasn’t looking for it, the reverse (low ROC) is also a fairly good indicator of market bottoms. But of course, this is just too small a sample to be robust. At best it is a good starting point.