Corporates won’t trade through Treasuries

By Felix Salmon
August 17, 2010 today:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

A rather startling pronouncement pops up on today:

“I wouldn’t be surprised if someday JNJ trades at a spread below Treasurys,” says Jack Ablin, Chief Investment Officer, Harris Private Bank. In other words, Johnson and Johnson yields could dip below the so-called risk-free yield on government debt.

Ablin might not be surprised, but I would be absolutely astonished. (Remember that Ablin is talking about Johnson & Johnson’s bonds, here, not the dividend yield on its stock.)

In the world of emerging markets, occasionally you’ll find corporates which trade “through the sovereign”. Generally, those companies are multinationals with strong and predictable dollar-denominated income streams, and the comparison is with dollar-denominated bonds issued by their home country.

It’s another thing entirely, however, for a company to trade at a lower yield than its home country can borrow in its own currency. And when it comes to the US dollars and Treasury bonds in particular, I’ll quite happily go out on a limb and say it’s simply not going to happen.

For one thing, if there’s a sell-off in Treasury bonds, that’s most likely going to be a result of worries about inflation — which eats away at Johnson & Johnson’s coupons just as much as it eats away at the US government’s. The only way for JNJ to trade through Treasuries would be if somehow there were worries that the US government was simply not going to pay some of its bonded obligations — while at the same time there were no worries at all about JNJ making its coupon payments. I can’t see it.

What’s more, Treasuries come with a substantial liquidity premium attached: to a first approximation, you can buy or sell them in any quantity, at any time, without moving the price. That’s very valuable — and it doesn’t apply to JNJ bonds. It’s also why Treasuries trade at lower yields than other securities without any credit risk, like World Bank bonds — which are guaranteed not only by the US but also by all the other World Bank shareholders.

So I’ll happily give Mr Ablin generous odds that JNJ bonds will never trade through Treasuries. Alternatively, I’ll buy some kind of financial instrument which pays off if JNJ never trades through Treasuries. The only problem is that no one would be willing to sell me such a thing.

Update: In the comments, tyler7 brings up a notorious Bloomberg article from March, which I dealt with at the time. People are always on the lookout for this kind of thing, but it never bears a thorough double-check.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Is the requirement that the debt be dollar denominated?
I bet that in the German hyperinflation there were many instance of German multinationals borrowing in pounds, francs, or dollars at lower rates than the German government could borrow in Deutchmarks.

Posted by OneEyedMan | Report as abusive

Felix … Berkshire has traded below Treasuries several times recently, and according to this article, J&J also traded below Treasuries back in March newsarchive&sid=aYUeBnitz7nU

Posted by tyler7 | Report as abusive

If you take state taxes into consideration, JNJ bonds might be close to Treasuries even now.

But I agree that inflation-driven default is more likely than formal default for the US.

Posted by TFF | Report as abusive

@tyler7, in the cases of this I am aware of, the spreads were tiny – less than 5bp.

At these differences, comparing nominal yields does not give an accurate indication of relative credit. The quotes are for actual rather than notional bonds, which differ slightly in maturity and nominal coupon. Also, quoted yields differ slightly from actual yields, as they do not account for date adjustment and only approximate accrued interest. Note that treasury yields are calculated ACT/ACT but corporates 30/360.

The only accurate way to measure such small differences is to produce the exact contractual cashflow dates and amounts and then discount them using a treasury zero curve that recovers the market price of your treasury. If the NPV of the corporate is greater than this, then it is trading at a positive implied credit spread. If it is less, then you have a story! But usually the story is just somebody covering a short or some other temporary technical demand factor.

Posted by Greycap | Report as abusive

OneEyedMan, yes, absolutely, the debt has to be dollar-denominated, that’s the whole point. Corporates can’t borrow through sovereigns who can print the currency in question.

The United States seems headed toward eventual default, with large deficits as far as the eye can see, deficits which worsen with time. It seems to me the debt has already past the point of where it be properly managed.

Everyone assumes that a US default is simply inflation, which has happened many times in our history and which is nothing to write home about.

Is there simply a zero chance of a haircut being applied to treasuries? Do we simply assign a 100% probability to inflation and a 0% probably to a haircut?

Posted by DanHess | Report as abusive

“Do we simply assign a 100% probability to inflation and a 0% probably to a haircut?”

If we are speaking in general, outside the US context, the answer is no. For instance, during its 1998 crisis, Russia defaulted on its short-term GKOs, domestically denominated, but honored many long term foreign currency obligations. The opposite of “the central bank can always pay in its own currency.”

But note the circumstances. The *reason* they defaulted on the GKOs was unaffordably high rates, and the reason for the high rates was to attract foreign currency; you had to buy rubles to buy GKOs, and Russia was trying to defend its exchange rate with high domestic rates. Once they gave up and floated, everyone wanted out of all ruble-denominated assets instantly. Government securities took a nasty haircut but corporate ones became smoking radioactive trash (please, no ruble/rubble jokes.) You don’t find a corporate rated more highly than its government in this situation.

Posted by Greycap | Report as abusive

Dan, any time somebody is 100% certain about ANYTHING, they are dead wrong. And I’m 100% certain of that.

But what would be the financial and political implications of a haircut? What would be the implications of 50% inflation over a five-year period?

In either case, it would become very costly for the government to borrow for a while, so there would be forced austerity. This would generate recessionary pressure, making inflation perhaps the preferable “solution” since there is nothing worse for a debt-ridden society than attempting to meet fixed obligations on a declining income. Moreover, the cost of borrowing following an inflationary haircut would rapidly ease once it was apparent that the government had its fiscal house in order.

However a haircut, executed sensibly, would have less impact on the private sector. Paradoxically, people might continue to trust the dollar while (justifiably) being fearful of the government’s willingness to pay its debts.

My recommendation would be to begin by balancing the budget over the next five years (and yes, that will essentially kill any “recovery” we might otherwise see). If it can be done while maintaining zero growth, rather than a massive drop in the GDP, then interest rates ought to remain low. Refinance long! At that point, steady but persistent inflation (3% to 5%) for a decade ought to cure many of the imbalances in the system.

Or maybe, if we get lucky and the austerity measures are wholly embraced, we can meet the obligations without default. After all, it isn’t like the debt service costs us anything right now.

Posted by TFF | Report as abusive

Back to the earlier question about the bond being dollar-denominated. Interesting question. I suppose the theory is that the default-by-inflation, resulting in USD devaluation, would cause such a shift that non-USD denominated multinational debt trades through Treasuries.

Just like the previous example of emerging market multinational corporate dollar-denominated debt trading through the sovereign, JNJ could float a global issuance in euro or Sterling or some currency basket and isolate the USD risk.

You could easily make a synthetic euro or Sterling security using existing JNJ and corresponding currency swaps, CDS the default risk away, add an inflation swap, and see where that gets you as far as yield goes. Probably pretty close to TIPS. Maybe below.

Where is the GBP-USD swap curve these days, and where is JNJ CDS trading? Wish I had Bloomberg Anywhere so I could find out . . .

Posted by 130-30 | Report as abusive

130-30, the multinationals already borrow in a variety of currencies. I think they are called “Samurai bonds” when they tap the Japanese market (denominated in yen). Euro bond issues are also pretty common.

And yes, the yields in Japan are very low, as the currency is perceived to be free of inflation risk.

Posted by TFF | Report as abusive