Pure Guesswork

Ten year yields were not falling last week. Which may seem surprising to anyone who was paying attention to the financial media last week. What we actually saw last week was an excellent example of the behavioral finance principle that the market's perception of the present (and future) is actually a perception of the past. When New York traders got to their desk last Monday morning the ten year was trading at 2.62%. When they left on Friday it was at 2.61%. This morning it is at 2.63%.

Rates had begun falling below their June and July lows on Friday, August 6, the day of the last employment report. They continued falling throughout the following week and last Monday's open was a gap down from the previous Friday close. But basically by end of day Monday we had seen all of the decline. The action during the week might be described as range bound, with sellers coming in above 2.65% and buyers around 2.55%. There was a tendency to probe to the downside, but none of these probes resulted in continuation.

We can set up a possible Fibonacci pattern on the daily chart with a base formed from the low in late May (around 3.10%) and the high in early June (around 3.42%). Projecting extended targets downward we get theoretical targets at the 1.618 and 2.618 levels. The 1.618 target was approximately 2.90%. And we find that this level formed a floor through all of June and July--indeed, up until the breakdown on the day of the employment report.

The 2.618 target would be around 2.57%. Did we see that level come into play last week? A look at the daily chart shows that while rates probed below that level, most activity remained above it.

This is not to say that it is time to call a bottom in rates. But it is likely that we will see at least sideways action and probes to the upside in the near future.

If the decline were to continue this same Fibonacci pattern would call for the next target at the 4.236 level (1.618 to the third power). That would take us to 2.04%

Posted by Pure Guesswork at 10:59 AM | Permalink | TrackBacks (0)

Playing around with some Fibonacci patterns on a weekly 10 Year Yield chart.

First, a pattern noticed during the impulsive move from March to June of 2009:

An alpha-beta base formed from the January low to the March low was followed by an impulsvie move up to approximately 4.00%. The 2.618 Fibonacci target of this pattern was 4.008%, the actual high at 4.014%--a 0.6 basis points difference between target and actual high.

The pattern since that time seems to be a 3 wave non-impulsive correction of the move in 2009 to 4.00%. If that is the case we are currently in the C wave down.

Some potential targets for this C wave:

Another:

Posted by Pure Guesswork at 10:16 AM | Permalink | TrackBacks (0)

A comparision of the rate differential between 2 Year US Treasuries and Bunds. From June 2008 until March 2009 the spread narrowed. As this happened the euro topped out, and then declined. When that trend ended and we had a relatively flat relationship between the yields of the two, the euro began to rally. The question is how long this can continue if the spread does not begin to increase again. The rate advantage of the 2 year Bund now is scarcely 25 basis points. The last time the euro surged above 1.50, the rate advantage was well over 100 basis points and rising.

Posted by Pure Guesswork at 12:03 PM | Permalink | TrackBacks (0)

Yield has fallen to first level of support, around 3.65%. This is approximately a 23.6% retracement of the impulsive move off the 2.46% low yield (touched in March). A 38.2% retracement would take us back to an approximate 3.40% yield.

Posted by Pure Guesswork at 3:35 PM | Permalink | TrackBacks (0)

A suggested extended Fibonacci pattern taking form in the 10 year treasury yield.

The base from which the extension is being formed consists of the 3.05% top made on February 9 and the 2.46% low made on March 19. Notice how the market flirted with this top several times through February and March, but it was only after a significant fall to the March 19 low that it was able to mount a rally that broke through this resistance. (Obviously what is called a "rally" here in rates, was a bear move in the actual treasuries themselves.) The move became impulsive once the 3.05% level was broken, never falling below the top of the base since that date.

The first level of resistance projected by this pattern would be at 1.618 times the base, or 3.41%. The market approached that level on May 8, topping off at 3.39%, selling off for several days and then finding support right above the 3.05% base top 5 trading days later.

The next significant target--and a possible end point of this pattern--would be at the 2.618 extension, right around 4.00%. Say 3.95% to 4.05%.

Above that, the 4.236 extended target would come in at 4.95%.

Posted by Pure Guesswork at 7:05 AM | Permalink | TrackBacks (0)

An Historic Day: September 8 2008

A brilliant analysis of what happened today in the bond market by John Jansen. The fall in treasuries, predicated on selling less risky bonds in order to buy riskier assets was reversed when the mortgage market loosened up, creating a demand for treasuries from mortgage servicers. As John Jansen says, "Always expect the unexpected is a fine mantra to follow even in the bond market."

Posted by Pure Guesswork at 8:45 PM | Permalink | TrackBacks (0)

U.S. treasuries were soaring last week as foreign accounts unloaded agency debt and replaced it with official government paper.
The move off the low in June happened in a 3 wave correction.
Weekly chart of the nearby 10 year future:

On a daily chart one can see how the third wave of the correction traded up to the 1.272 fibonacci level, then found some resistance. Then on Tuesday of last week it broke through this level and moved swiftly up as the switch from agencies to treasuries accelerated, bringing it on Friday to the 1.618 level. The news on Friday afternoon that the bail out of the GSEs would occur over the weekend (as it has) broke the rally, and the market fell back to the 1.272 level (making Friday a key reversal day).

The implicit guarantee of GSE bonds is now explicit. The result: this afternoon the national debt of the United States of America increased by 5,400 billion dollars. There is now no good reason to switch from agencies to treasuries. Bunds, anyone?

Posted by Pure Guesswork at 4:45 PM | Permalink | TrackBacks (0)

Posted by Pure Guesswork at 8:37 AM | Permalink | TrackBacks (0)

A 4.236 delta target was met on the move from last June's low to the high in March.

Since the high (made on the Sunday night of the announcement of the Bear Stern sales to JP Morgan Chase), it has retraced about 50% of the impulsive move from beta (in mid-October 2007) to the extended panic high. An irregular X-Y-Z pattern can be seen, where Z = X.

So for now the bottom might be in around 114'20 (note charts are not in 32nds, but decimal) on the June contract, corresponding to about 3.90% yield on the cash bond.

Posted by Pure Guesswork at 6:42 AM | Permalink | TrackBacks (0)

Updating the charts posted on Friday and Thursday, a minor bull move in rates (bear in bonds) met some resistance at the 3.90% level on Friday, a 2.618 delta target. But both downtrend lines have clearly been broken. (See the posts referred to above for weekly charts showing how these trendlines were developed.)

Posted by Pure Guesswork at 9:51 AM | Permalink | TrackBacks (0)

First the nearby 10 year future on a weekly chart:

Then an updated weekly and daily chart of the TNX CBOE index of 10 year rates. (Note: I have put in an overnight high above 3.90%, although the official index only marks prices made within traditional NY 8:00 AM to 3:00 PM bond market hours. This is one reason TNX charts and CBOT 24 hour futures charts diverge at times.)


Resistance on rate chart is around 3.91% (high in February was 3.96%). On futures chart, resistance in June futures could come in at 114'12. The low in February was 113'25. Overnight future traded down to 114'20. It has rallied significantly (18/32nds) since then.

Posted by Pure Guesswork at 9:07 AM | Permalink | TrackBacks (0)

Two trendlines. The shorter term line is steeper, was tested in February and broken last week. The longer term line, dating from the top last June, has not yet been broken. Market traded up to it last week (3.85%) then bounced down.

And then plotting these trendlines on a daily chart:

Posted by Pure Guesswork at 9:10 AM | Permalink | TrackBacks (0)

Matching a 2.618 delta downside level on the S&P yesterday:

On the nose.

Posted by Pure Guesswork at 10:15 AM | Permalink | TrackBacks (0)

On DJ Newswire this morning there is a quote from a fixed income analyst saying, "Euribors are getting numb to an ECB that barks but refuses to bite. The chart of the June Euribor future below demonstrates this. A 30 basis point move since Friday.

And then the September future:

Posted by Pure Guesswork at 9:55 AM | Permalink | TrackBacks (0)

If the pattern on this chart is valid, the 10 year treasury is rapidly approaching a 2.618 delta target. Yesterday afternoon (January 15) it reached 3.70%. This morning at 6:00 AM EST it is trading at 3.64%. The 2 year, which has a pattern with a 2.25% goal, is at 2.46%. Libor rates, which very recently were trading an average of 60-75 basis points above FF rates, are now trading below current FF rates, approximately 100 basis points lower than they were a month ago.

Posted by Pure Guesswork at 5:55 AM | Permalink | TrackBacks (0)

John Percival notes that while these indices dipped again last week, they failed to make new lows. Does this presage a return to a coming return to a more robust risk appetite in the markets?
AAA:

BBB:

Posted by Pure Guesswork at 9:41 AM | Permalink | TrackBacks (0)