What Did Bernanke Say?

Answer:  Not very much, but the market reacted anyway.

The Fed Chairman testified before the Joint Economic Committee of Congress today in Washington, and the message was the same as before:  If the economic data indicate a sustained labor market recovery, the Fed may step down asset purchases after the next few Fed meetings.  Bernanke emphasized, however, that ending QE3 prematurely would endanger the recovery.  The Fed is going to let the data dictate their actions, which is what Bernanke and the Fed governors have been saying all along.

The market, however, interpreted this as a higher likelihood of QE3 ending sooner than anticipated, and interest rates soared.  The 10-year yield jumped 11bps today, from 1.93% to 2.04%:

Yields 2013-05-22

Although I believe that yields will move substantially higher over the next few years (remember, the blog’s called “Higher Yield”), I think the market is oversold in the short term.  We’ve had a big run-up in yields over the last three weeks, about 40bps on the 10-year Treasury:

Yields May 2013

 

If the weekly initial unemployment claims disappoint tomorrow morning, I’d expect a big drop in yields.  As readers of this blog know, on a monthly basis I’m focused on the weekly continuing claims as a predictor of the monthly new jobs number.  But initial claims are a stronger driver of short-term yield movements, since they’re the focus of most market participants.  The forecast (according to Bloomberg) is for initial jobless claims of 345K.   If we get an initial claims number 350K or above, I think there will be a rally in bonds and a drop in yields.

How to trade this?  One could buy treasuries or treasury futures, but purchasing calls on treasury futures looks more interesting.  A risky play is the USM3 (June treasury bond futures) calls with strike 144 that expire Friday.  With the underlying security (USM3) at 142 17/32, the last trade in these calls was 5/64. The 10-year yield would have to fall about 12bps or so for the treasury future (USM3) to rise to 144 and then the option would be just at the money.  If that happens (say, if the initial claims come in at 360K), I estimate the call will trade around 25/64 or 30/64, a nice gain.  However, if the initial claims are reasonable or good, say below 350K, yields will stay the same or rise, the price of the calls will drop, and it may be tough selling out of the position for a price greater than 1/64 or 2/64, meaning a 60% or 80% loss.

A less risky trade is buying USU3 (September treasury bond futures) calls that expire June 21.  The 145 calls are trading about 28/64, and a big drop from yesterday’s close of 50/64.  If yields come back down to where they closed yesterday on a bad initial claims number, it’s reasonable to assume these calls will revert (near) to yesterday’s price of 50/64 – the loss in time-value from being closer to expiry will be (roughly) cancelled out by the increase in volatility.  If yields stay where they are after the claims announcement, I wouldn’t expect the option price to change much.  But if we get a great jobs number, say 315K or better, yields may continue their march upward, and these calls could also lose a large chunk of their value.

PLEASE NOTE:  Nothing in this blog post or other blog posts should be taken as trade recommendations or advice, and you should not invest based on anything you read in this blog.  Every investor’s situation is different, and specific trade ideas are never appropriate for every individual investor.

Woulda, Coulda, Shoulda

I really felt last night that it would be a great time to lighten up on US Treasury short positions ahead of the economic news this morning.  We had a big run-up in yields in a short amount of time, from 1.63% (10-year U.S. Treasury Yield) on May 1 to 1.93% at Wednesday’s close, and the market seemed ripe for a pull back if the data didn’t meet expectations.  Well, if you followed what happened this morning, the weekly unemployment claims, consumer price index, housing starts, and the Philadelphia Fed Business Outlook Survey all came in worse than expected.  Bonds rallied, and the 10-year yield dropped from 1.95% before the claims and CPI announcements to 1.90% after, and closed the day at 1.88%.

The fact that a particular investment move turns out to have a good result doesn’t mean it was a good idea – someone’s going to win the lottery, but that doesn’t mean you should throw your life savings into Powerball tickets (though there are times when a lottery can have a positive expected value). But this was a good idea. Given that I believed that the magnitude of the potential drop in yields given bad news was greater than the magnitude of the potential rise on good news, and that this morning’s news was slightly more likely to disappoint than surprise given inflated expectations, it would have been the right move to lock in some gains at 1.95%.  However, I didn’t follow through, and you don’t get any credit in investment management for having a good idea and not executing it.

So why didn’t I pull the trigger? Mainly because I didn’t do a good enough job convincing myself that it made sense to reduce short positions – after all, if the news was good this morning, yields would have risen further. Before taking action, I needed to do more analysis on the economic data forecasts, the likelihood of beating those forecasts, and the subsequent impact on yields. Even if the probabilities are subjective, this type of detailed analysis helps clarify my thoughts and makes the investment decision much clearer.

There’s no point in compounding mistakes, so I cut positions in half right after the first downdraft in yields, locking in some gains.  My next decision is whether to cut positions in half again.  The argument in favor of a further cut in positions is that the economic news today was lousy, there will not be much other economic news for a week (so little chance for positive surprises), and yields may be vulnerable to an equity market sell-off.  The argument for keeping a short position is my continued belief that yields will rise. This time, I’ll be sure to do a more thorough analysis to support my decision.

 

Big Move in JGB Yields

What a move in Japanese Government Bond yields over the last three days!  The chart below shows the yield on 10-year JGBs since the beginning of March, which had been holding steady around 0.60%, but just spiked up to 0.86%:

JGB Mar - May 2013

Here’s a longer (2-year) view of the same 10-year JGB yields.  Note we’ve blown through the 50, 100, and 200 day moving averages – next resistance at 1.00%?

JGB 2011-13

A 26bps move in the last three days was a welcome sight for people short these JGBs (see here for investment vehicles to implement this), providing about a 2% gain.  The move, however, goes against Bank of Japan Governor Kuroda’s asset purchase goal of keeping the long end of the yield curve low (Japan’s version of quantitative easing).  Why such a large move up in yields if Japan’s central bank is buying substantial amounts of bonds?

Here are four possible explanations:

1.   Over the past year, the Japanese inflation rate has averaged about -0.3%.  That means a small amount of deflation, though the magnitude has been larger recently (-0.9% CPI for March).  If nominal yields are 0.6% (on 10-year JGBs) as they were recently, that means a real yield of 0.9% based on recent inflation values.*  The Japanese central bank recently announced an inflation target of 2%; combined with the 0.6% nominal yield, this produces an implied real yield of -1.4%, a very unappealing number.  For those investors focused on real yield,  the nominal yield will need to rise to make these bonds attractive.  See the chart below for real yields in Japan over the past 20 years (bottom left panel); they have rarely been negative.

.Japan Real Yield

2.   Japanese equities have gone gangbusters this year, up 42% YTD.  This is partially in response to the devaluation of the yen, which has fallen from 80 to the dollar in November to 102 today.  Expectations are that Japanese corporations will be able to profit handsomely from the weaker yen, making equities much more appealing than JGBs with a negative real yield.

3.   The role of the JGB as a safe-haven asset may be compromised.  (As an example, JGB yields fell 20bps in the summer of 2011 as the European crisis flared up, and back in 2008 they fell 70bps during the global meltdown – both clear safe-haven indications.)  But a foreign investor now might not view the JGB as safe if losses will mount due to currency devaluation, or hedging the currency risk is viewed as too expensive.  Better to invest in a “safe” country that is not trying to depreciate its currency.  Some data to back this up:  Bloomberg reports that in March, overseas investors cut their JGB holdings by the largest amount in three years (no April data yet to show whether this divestiture has continued).

4.   On the same devaluation note, Japanese investors may want to diversify into overseas assets as the yen depreciates.  Below is a chart of the yen in dollar terms over the last 40 years.  It’s reasonable to think that the yen could depreciate another 20-30% from here and still be on the low side of the historical range.  I’ll be interested to see the flow of funds data for Japanese investors to verify whether this is the case.

JPY 1973-2013

Although this was a big move in a short time, and I expect volatility in JGB yields in both directions, I think the short JGB trade still has room to run.  Don’t forget that Japan’s government debt is 250% of GDP, the largest in the world.  This is a longer-term issue for the government, but sometimes longer-term issues get accelerated to the near-term by markets.

—–
*Note that we are mixing apples and oranges to some extent, since the JGB is a 10-year yield (long term) and the CPI is a recent short-term value.  Ideally, we’d like to calculate real yields based on longer term inflation expectations.  However, one could view Kuroda’s 2% inflation target as a longer-term expectation

Mixed News, As Usual

Some good economic news today and some not-so-good news.  The weekly unemployment claims came in at 323,000, the lowest number since January of 2008.  This is an encouraging statistic, but, as Higher Yield readers know, I’m more interested in the continuing claims. This result was also encouraging, falling by 27,000 to 3,005,000 for the week ending April 27, in a continuation of the steady downward trend from the peak in mid-2009. If continuing claims fall under 3,000,000, the May jobs number release in the first week of June should be good, so pay attention to this number over the next few weeks.

Continuing Claims

Continuing Claims (Source: Bloomberg)

Mortgage news was mixed today. The Mortgage Bankers Association’s quarterly report showed an increase of 16bps in the delinquency rate (one payment past due) from the prior quarter to 7.25%.  However, on a non-seasonally adjusted basis, the rate was down 76bps from last quarter and down 19bps from the same quarter a year ago.  The serious delinquency rate (90 days or more past due or in foreclosure) was down 39bps from last quarter to 6.39% and down 105bps from a year ago.  I view this as mostly favorable, though the uptick in the delinquency rate shows that there are still some borrowers having problems making their payments.

On the negative side, the Merchant Wholesalers Sales Total Monthly % Change came out at -1.6% versus a forecast of +0.1%.  Clearly a bad number, but keep in mind two mitigating factors:  First, there is a 6 week delay in this data, so this monthly change is from March, which we know to be a weak month from other economic measures.  Second, this series jumps around a fair amount even in good economic times and with seasonal adjustments:

Wholesalers Sales

Monthly Change in Wholesalers Sales (Source: Bloomberg)

Retail sales for April come out on Monday, which should give us a more recent look at the sales picture.

Putting all of these reports together, I would say the positives (jobs) outweigh the negatives in mortgage delinquencies and wholesale sales. Unless I see more negative economic news, I still believe yields will head higher.

Technical Note

The yield on 10-year U.S. Treasury notes was up about 4.5bps today according to Bloomberg, from 1.766% to 1.811% — a fairly large one day increase.  Yahoo Finance and CNN Money had a similar rise, from 1.76% to 1.81%.  You might have expected the 10-year Treasury futures contract (June expiry) and the Ishares 7-10 year Treasury ETF (ticker IEF) to be down substantially, but the futures contract closed only down 3/32, or -0.05%, and IEF was only down 0.02%.  The reason for this is that the generic yield for 10-year U.S. Treasuries comes from the most recently auctioned bonds.  Since there was a 10-year auction yesterday, today’s rate quote was from a different bond than yesterday.  Specifically, the yield had been based on the 10-year note expiring Feb 15, 2023 first auctioned in February, and now it will be based on the 10-year note expiring May 15, 2023.

Note that both the old (Feb) note and the new (May) note were essentially flat today.  I think it’s misleading for the news services to report the yield as up 5bps when it actually hasn’t changed – it’s just that the yield on the May note is slightly higher than the February note since it has a longer time until expiry.   This is essentially the same problem as reporting changes in futures prices when rolling to the next month’s contract.  I don’t know of any easy solution – it’s hard to report a single data series (i.e., a single data item) for each day and incorporate all available information.  Keep this in mind when looking at yield changes the day after any new government bond auction.

Where’s the Unemployment Rate Headed?

The April jobs numbers released on Friday were very good.  The Establishment Survey showed a gain of 165,000 in Nonfarm Payrolls – a bit less than I was expecting, but more than the forecast of 140,000.  In addition, the March gains were revised up by 50,000 jobs and the February gains were increased by 64,000.  The Household Survey showed the unemployment rate falling from 7.6% to 7.5% — another positive sign.

Markets reacted strongly.  The 10-year U.S. Treasury Yield gained 11 bps, from 1.63% to 1.74%.  For those who read my last post, the June U.S. Treasury puts with a strike of 145 started the day at 8/64, reached an intraday high of 27/64, and closed at 22/64 for a 175% gain.

Unemployment

The Fed has made the unemployment rate a crucial policy indicator by linking it to future interest rate hikes – the threshold for raising rates is an unemployment rate of 6.5%.  In addition, the Fed has indicated that before it raises interest rates, it will stop Treasury purchases (QE3), which will clearly be bullish for yields.  Since this makes unemployment a more explicit driver of yields, it’s worthwhile for us to look at where the unemployment rate might be headed.

The unemployment rate is simply one minus the employment rate, which is the number of employed divided by the civilian labor force.  Focusing on the denominator of this fraction first, here’s a chart of the civilian labor force since 2012 with a linear trendline:

Labor Force

The labor force has been increasing by about 90,000 per month.  Note that this is less than the overall increase in civilian population, which has been about 200K per month.  This means that the labor force participation rate has gone down over this time period, from 64.3% to 63.3%.  This is mainly due to increased college enrollment and graying baby boomers dropping out of the workforce – see this post from Calculated Risk for more details.

The numerator is the number of employed which has been trending up by about 180,000 per month since 2010.  If we assume this continues (180,000 new jobs per month is a reasonable assumption for a modestly growing economy), and that the labor force keeps increasing by 90,000 each month, unemployment will fall by about 0.06% each month.  By the end of this year, the unemployment rate would drop from 7.5% to 7.0%:

Unemployment Rate Forecast

For those that like sensitivity analysis, if jobs increase by only 120,000 per month the unemployment rate will be 7.3% at year’s end, and if jobs increase by a more robust 240,000 per month, the unemployment rate will be 6.7%.  Holding jobs increases at 180,000 per month, if the labor force only increases by 30K instead of 90K, the unemployment rate will end the year at 6.7%, and if the labor force increases by 150K instead of 90K, the rate will be 7.3%.

There are lots of “ifs” in this analysis, but if you believe the Fed will stop Treasury purchases if the unemployment rate is 7% and falling, it’s reasonable to think the Fed will stop the program sometime near the end of the year, which should push yields higher.

April Jobs Report Tomorrow

It’s hard to believe that since the start of February, the S&P 500 Index is up 100 points, almost 7%, while the yield on 10-year U.S. Treasury Bonds is down 36bps, from 1.98% to 1.62%.  Needless to say, this hasn’t been a good time to be short Treasuries, but the market may start turning tomorrow with the monthly jobs report release.

Here’s a chart I’ve shown before with the relationship between the monthly change in non-farm payrolls (vertical axis) and the change in continuing claims from the weekly unemployment claims report (horizontal axis).  To calculate this change in continuing claims, I took the average from the current month minus the average from the prior month.    The six times since the beginning of 2011 we’ve had a 60,000 or more reduction in claims, the monthly change in payrolls has been 200,000 or higher, a very good number.

April Jobs

Let’s look at the change in average continuing claims from March to April to help forecast what tomorrow’s non-farm payroll numbers will be. The average continuing claims number for March was 3,087,000, and for April (through the 19th) continuing claims are 3,013,000 – a reduction of 74,000.  Thus I wouldn’t be surprised if tomorrow, we see an increase in non-farm payroll of more than 200,000. Since this would be significantly higher than the expected number of 140,000 (based on the Bloomberg forecast), I expect this news would push up yields.

From a trading perspective, I like the June puts on the long bond futures (Bloomberg ticker USM3P) with a 145 strike.  With a price of 8/64, they should have good upside in case of a yield spike, and I expect to be able to exit the position with a small loss at 7/64 or 6/64 if the jobs number comes in around the forecast of 140,000.

Weekly Claims and Monthly Jobs

I’ve written before about the relationship between the number of new jobs created (change in non-farm payrolls) and the unemployment rate, both of which come out once per month (usually on the first Friday of the month).  A more frequently released item is the jobless claims data, which comes out weekly on Thursday.  The series that gets the most press from the weekly release is the Initial Jobless Claims, the number of people filing for unemployment benefits for the first time.  This weekly series is notoriously volatile, but can it give us any indication of what the monthly data will show regarding the number of new jobs created?

You would expect that the fewer the initial unemployment claims (weekly), the greater the number of new jobs created (monthly).  Unfortunately, there doesn’t seem to be a statistical relationship.  The chart below shows the average weekly initial jobless claims on the horizontal axis and the number of new jobs created for the same month. Each point represents a particular month in the period from January 2011 through February 2013.

Jobs vs UnemploymentIf the outlier in the upper right corner of the chart were removed, you might be able to argue that there is a small, negative relationship between the two series, but I don’t find it convincing.

We see a more interesting result if we look at a second series from the weekly release, the Continuing Claims series, which is the total number of people currently receiving unemployment benefits. I calculated the monthly change in continuing claims by taking the average value for the current month minus the average value for the prior month. This value is on the horizontal axis in the chart below, while the vertical axis again shows the number of new jobs created. The data period is the same as in the first chart, from January 2011 through February 2013.

Jobs vs Continuing

This seems like a much stronger relationship.  In the seven months since January 2011 when the continuing claims fell by 50,000 or more, the number of new jobs created has been very good (195,000 or more).  Based on this brief analysis, I would pay more attention to the continuing jobless claims than the initial jobless claims. In February, the average continuing jobless claims were about 3,100,000, so if the March average for continuing claims falls by 50,000 or more (to 3,050,000 or less), the jobs number released in the first week of April should be good.

————————————————————————————–

Cyprus update:  No news is presumably good news as far as Cyprus is concerned.  U.S. equity markets were up today as were yields on U.S. Government Bonds.  The president of Cyprus is supposed to present a revised plan to Parliament tomorrow.