The interest rate riddle

The US current account deficit came in yesterday at $164.7 billion for the third quarter of 2004, lower than expected but still a new record. The previous day saw the trade balance for October, also a record deficit. As usual General Glut has detailed coverage.

In the face of all this, long-term interest rates, as measured by the yield on 10-year Treasury bonds, are falling. The rate is currently about 4.2 per cent. By looking at inflation-protected bonds (TIPS) it’s possible to work out that this is made up of an expected CPI inflation rate of 2.5 per cent and a real interest rate of 1.7 per cent. As the Fed has increased short-term rates, the margin between long and short rates is falling, and seems likely to go close to zero.

This makes no sense at all to me. Given the near-certainty of further depreciation of the $US in the long run, who would buy 10-year bonds at rates like this, who would hold them if they had them, and why doesn’t someone like Soros short them? The answer to the first question appears to be “non-US central banks”, the answer to the second must have something to do with institutional inertia. As regards the third, and relying on introspection, the answer may be that the market can remain irrational longer than Soros can remain solvent (certainly that explains my non-participation). Soros took a hammering, if I recall correctly, betting against the NASDAQ in 1998, and the fact that he was proved right in the end is cold comfort.

Since I don’t believe that capital markets are efficient or collectively rational in the short or medium term, this kind of thing doesn’t pose a fundamental problem for my worldview. Still, I can never quite stop being surprised when asset prices are so obviously wrong.

28 thoughts on “The interest rate riddle

  1. The US bond market could plunge today, tomorrow or in 2 year’s time.

    It’s like that blonde woman says on the shampoo commercial: “it might not happen right away, but it will happen”.

  2. JQ,

    Perhaps you should dabble in the markets – if only for your further education. You’re not a lonely voice in the wilderness on this issue.

    Institutions ARE shorting the USD, and ARE bailing out of US treasuries. Warren Buffett of Berkshire Hathaway announced he was doing so a year ago.

    You say that the pricing of US 10-year bonds is irrational, but for whom? Asian central banks have been guzzling US paper for years now, and continue to do so. Why? Not because of the miserable yield, as you rightly point out, but because of the dampening effect it has on their currency relative to the USD. China, Hong Kong and Singapore have pegged their currencies to the USD, and other countries in Asia (notably Japan) actively manage their exchange rate through the trading of low (credit) risk US treasuries. Allowing their currencies to appreciate relative to the USD would have significantly deleterious consequences for their export sector, which are highly dependent on US consumers importing their goods. They are thus artificially supporting the USD.

    I think it’s unsustainable, but the Asian central banks don’t seem to have thought out an alternative. This means the USD will have further and further to fall, and US yields that much higher to rise, when the whole thing unwinds. It’s going to be a very rocky ride.

  3. I am sure your argument “the market can remain irrational longer than Soros can remain solvent” is the reason that no-one is shorting 10 year bonds, yet.

    During the recent stock market bubble it did not take a genius to realise that prices were way too high, but a lot of very smart investors learnt the hard way that just because prices are too high does not mean you know when they are going to come down.

    If the stock market bubble could keep going so much longer than everyone expected it to with participants who were only interested in the returns they were getting, how much longer can the bond market stear clear of reality when its mainly being driven by government central banks who have truly enormous funds available to them and are motivated by exchange rates rather than returns?

    I am sure every smart investor can see that the music has to stop at some point but, but as precarious as the current situation is, they believe that it is possible for it to be maintained for a substantially longer period of time. That might sound unlikely, but try constructing a solid argument that the coming re-adjustment has to happen in the first half of next year, or the second half, or any other specific time and you will start to see their problem.

    They have decided instead to wait for until the bond market actually starts to move because as long as they react quickly enough there will still be plenty of money to be made without the risk of shorting a bond market that might not actually do anything for years.

  4. Fyodor, I agree that institutions are shorting the USD, but why aren’t they shorting US bonds as well ? Do they judge that they can’t outlast the Asian central banks, for example?

  5. I have lost some money lately being short 10-yr+ bonds. My perception is that the other side of the trade (longs) is more rational than you give it credit for. Whether yields go up or down really depends on what sort of growth happens in the future. The 10-yr could be saying the Fed is overestimating the economy’s strength (inflation potential) by keeping ON rates at 2.25%. I have tried shorting because I think the curve should steepen as people sell the long end to buy the front but so far that has not happened.

    Also, don’t underestimate the power of the money illusion. Lots of people in the U.S. only look at absolute returns in dollars and until the dollar fall starts causing a substantially higher CPI they will continue to do so.

  6. John,

    Saying the markets aren’t rational in the short or even medium term, is true in terms of the prices not reflecting fundamentals. It doesn’t mean that for individuals that buying at this level isn’t rational.

    Having worked in a dealing room or two the guys trading are interested in at most the current years P&L, because that drives their bonus, but mostly on the current day P&L. They trade on what they think everyone else is going to do and so most put much more reliance on technical factors as an indicator of what everyone else is doing. Fundamental factors are of interest in that they want to think about how the market (others like themselves) will react short term. Longer term fundamental drivers like what you mention are of much less concern.

    Its much better to lose with everyone else when the market crashes, than to lose on your own betting against the current wisdom. That is much more likely to lose you your job.

    In the end for these people, it is rational to be buying at these levels even if from a longer term investment point of view it probably not. For traders at least, that’s not their timescale. In addition their personal payoff is not just proportional to the rate of return they get from transactions, even if it is for their employer.

  7. JQ,

    Many are – US 10yr bonds were sold off 11bps last night, and at a yield of 4.19% they’re well off their recent lows of under 4.00% in October.

    I suspect two complicating factors for those going short are:

    a) the wall of money flowing in one direction, from Asian central banks to the US Treasury, which seems unlikely to stop so long as the US maintains a sizeable deficit in its current account with these countries; and

    b) concerns over the sustainability of the US economic recovery. If the US economy rolls over next year, current expectations of further Fed rate increases will be overturned, with short-term yields trending down. That would be a positive environment for long bonds, and negative for those shorting them.

    Personally, I think both the USD and long bond (price, not yield) will fall in 2005, as I believe Japan and China will agree to a step-change appreciation in their currencies rather than continue to over-stretch the USD and risk an outright collapse.

    The longer this goes on, the more powerful the correction. We need only note the GBP’s withdrawal from the ERM in 1992 to remember the power of market forces to humble interfering governments.

  8. There is no riddle if interest rate expectations drive exchange rates, rather than the other way around. I would suggest that exchange rates are a relatively minor factor in the determination of domestic interest rates, but expected interest rate differentials are a big factor in driving exchange rates. If the USD can wear most of the adjustment, that takes the adjustment burden off other markets.

  9. Fyodor,

    Do you have data on how the East Asian central banks are maintaining their portfolios? I’ve heard anecdotal evidence that they are investing in other currencies, well aware that they cannot prop the USD forever (and thereby maintain their trading advantage) – eg euro and gold, but I have little idea of the relativities. Also, as the Chinese (in particular) domestic market grows, dependence on exporting will lessen, and the need to maintain the USD will lessen. Surely that will compound the problem for the US long term.


  10. pwe,

    I don’t have that data to hand, so I can only suggest a couple of sources that might have it.

    There’s a lot of good data on foreigners’ purchases and holdings of US treasury paper at

    For each country, check out their central bank websites by googling Bank of Japan, People’s Bank of China, Hong Kong Monetary Authority, Monetary Authority of Singapore and Bank of Korea. Unfortunately, the data is often tricky to access and manipulate, but these websites are a good start.

    How do they maintain these flows? Typically by issuing their own paper in domestic currency and using the funds raised to invest in USD-denominated treasury paper. Remember that these countries have large current account surpluses with the USA and are generating significant savings for export into the capital-hungry USA.

    How long can this last? Hard to say, as it depends on their willingness to prop up the USD and the ability of markets to challenge their control of their exchange rate. The size of the reserves held by the Asian central banks suggests no private speculator could take them on, but they could decide at some stage to stop ploughing money into a depreciating asset (i.e. the USD). The size of their current purchases of US treasuries suggests they may be playing a risky game by doggedly fixing their currencies’ value to the USD. It’s also destabilising to the likes of the EU, which has now seen the euro appreciate considerably relative to the Asian currencies still pegged to the USD.

    On China, I think you’re right about the growth of domestic demand reducing the reliance of the Chinese on their export sector, but it’s unlikely to grow fast enough in the short term to be material to the exchange rate in the next 12 months.

  11. Why does the behaviour of the asian central banks remind me of that Gleeson guy (rogue trader)?

    Can somebody tell me exactly how the asian governments are financing their silly investments? Are they bleeding their efficient local industries or are they building up an inflation problem?

  12. In the early seventies, baby boomers were reaching new household formation age in unprecedented numbers and the Western world was racked quite unexpectedly by stagflation for a number of years. Now that same generation is reaching its fifties as empty nesters with much discretionary income and we notice some puzzling aspects of US interest rates, etc. Perhaps demography has some answers.

  13. In terms of the inflation/real interest rate decomposition, which is the unrealistic prediction?

    A falling dollar would be inflationary I guess but there doesn’t seem a lot of other presure. Is the real interest rate wrong? In the gloomy collapsing dollar scenario it would be hard to fit a shiny and strong US economy and the real interest rate will perhaps bear some relationship to per capita gdp growth. Finally if the Chinese, Indians and others do liquidate a substantial portion of their dollar portfolios the world will on the one hand be awash with bonds but on the other maybe also with cash so competition for access to dollar financing is unlikely to be severe. Possibly this is related to the shrinking corporate bond spreads.

    The rogue trader was Nick Leeson.

  14. I think that there might be an Armageddon U curve in here.

    A simplification of JQ’s thesis is that the US foreign debt is unsustainable if the US is to retain it self as the international currency with all its attendant benefits. Therefore as long as things continue as they are the risk/devaluation premium on US debt should rise.

    However its worth considering how things change depending on how far we are into “financial Armageddon”. Dividing the world up into Asia, Europe and the US there are 3 scenarios
    1) Crash in the dollar now – US looses reserve currency status and attendant benefits, Euro gains them and Asia gets a bit of a hit on debt looses and currency appreciations. Outcome/premium Rankings E > A> US
    2) Asian banks continue buying for 2 more years then sell, causing dollar crash. However under this scenario, the over valued Euro zone slumps economically and so can’t credibly replace the dollar as their economy is too weak. Similarly, the looses suffered by the Asian banks are so massive that a) it tips their economies b) they can not credibly back an international currency. US retains status as reserve currency. Rankings US>A>E.
    3) Complete breakdown in world financial and trade system. Rankings US > E> A as Asia is heavily dependent on exports and imported commodities, Europe and the US, less so.

    Therefore, if the Asian banks can credibly postpone the train wreck for another couple of years, the path of the risk premium passes from the US to the Euro and thence to Asia.

  15. Exchange rates and Asian Central bank intentions aside, you would hold or buy long term bonds because you expect returns on them to fall over their life, or any alternative returns are likely to fall also. IMO this seems more than reasonable at least until the end of the decade. At present, baby boomers and their super funds are progressively driving down the returns of all asset classes. After having done over the Sydney RE market they are driving the stock exchange index over the 4000 barrier and beyond in Oz. Finding satisfactory yields is becoming increasingly problematic and I anticipate this will continue apace until the end of this decade, when the first of the post-war boomers reaches the traditional retiring age of 65 and begins to cash out. In 2 or 3 years time, an interest rate return of 4.2% might be very attractive compared to today. Just ask the Japanese with their 1.4% return on govt paper.

    This sort of scenario begs the question as to what asset class would produce the best return over the medium longer term. IMO there would be 2 major classes- resouces and well watered agricultural land. Resource stocks are easily accessed at present, but there may be a lucrative new asset class with the latter. Investment companies, with capital raisings invested in well watered agric land and leased to farmers, with the shares traded on the stock exchange. The share price would reflect the undelying value of the long term assets over time, while allowing short term participation in ownership.

  16. Asians have traditionally been net savers, preferring to delay enjoyment until later, and also to cover for the bad times, since there is not any government welfare to rely on.

    However, these attitudes change. In 15 years time, the young Chinese will be more confident after three decades of 8% growth, and will be spenders. When this happens, there will be real competition for the products that are coming out of the country, and only then would the US have to work a lot harder to buy the goods.

    In the meantime, China is taking IOUs from the US because it’s financing the capacity building that is going on in the country.

  17. Giles, in your second scenario, what will change the policy choices and political incentives that created the situation in the first place?

    Does anyone have any ideas about whether it is inflation being underestimated or is it real interest rates? I’m not sure that there is obviously very strong reason for bonds to fall massively beyond what the dollar does. The yen is still way down on where it was ten years ago and Japanese interest rates have not exactly soared. Even a very weak dollar would encounter stiff resistance to inflationary pressure. The point of my question is where would the effect take place? Real rates or inflation? Both look farely tame at the moment.

    The far eastern example is very interesting. To some extent the high saving rates fill a gap left by small governments so they may give us a clue about what benefits might or might not accrue to a privatised social security system.

  18. If the ECB starts buying Yuan, it can make sure that scenario one occurs – since if the ECB goes against the weak Yuan policy, I can’t see how it can be sustained.

    On the other hand I cant see the ECB crossing sorwds with China, so its not likely,

  19. Jack,

    The Chinese realize that the way for their country to get ahead is by proving to other countries that they are capable of producing competitive, and high quality goods. This technology transfer doesn’t come cheap. The Chinese have effectively been paying for it by keeping their currency low, in order to fuel foreign investments into their factories.

    This is no different to apprenticeships where the apprentice accepts low wages initially in order to learn some skills, and also for the opportunity to prove that they are capable. Another example is how unemployed people offer to do volunteer work in order to ‘get their name out there’.

    Just a few years ago, goods coming out from China were regarded as inferior in quality, and look at how much has changed. While they are no where near Japan, Chinese products have crossed the line from being not good enough to being good enough.

    It is going to be a while before the Chinese is able to move up the value chain from simple outsourcing to serious R&D. And China cannot move up until their economy is rich enough.

    So how much longer would the Chinese keep buying US bonds? For quite a while I think. One needs to appreciate how tough their lives have been in order to understand how prepared these people are to sacrifice economically for the benefit of the next generation.

    The next generation of Chinese would be very different though. Since the economic growth plus the one-child policy will result in a youth who are more confident of their economic future. The younger lot will not have to save as hard as their parents, and will be more prepared to spend money.

  20. Being a lowly finance masters student I have been reading this commentary with great interest!!! And I was wondering if anyone could help me with a slight conundrum I’m faced with in reslts of regressions I’ve been running for an econometrics project. I’ve been testing the sensitivity of sectors within the japanese stock exchange (as a sample I chose non-ferrous metals and machinery) to changes in US interest rates, measured by the yield on the US 10 year bond and US 3 month t-bill. The 3 month was only added as an afterthought on advice from my lecturer and yet I’ve found that in both regressions I’ve been forced to drop the 10 year bond variable as it appeared not to be relevant! So I’m basically wondering why such sectors would be more sensitive to short-term, rather than long-term interest rates? Thanks.

  21. In your recent article, “The Unsustainability of U.S. Trade Deficits”, (, you posed the question, “Why Does the Dollar Remain High and U.S. Interest Rates Continue to Be Low?”

    I don’t believe you confidently answered this question. Do you know the answer to your question? It is a good question, and I’d like to know the answer to it.


  22. Kate,

    Over what time frame are you testing your regression, and how are you incorporating the US interest rates? That is, are you regressing the performance of the metals & machinery index relative to the absolute level of US 3-month and 10-year interest rates, or to proportional changes in these interest rates? On a real or nominal basis?

    It’s interesting/odd that you are analysing the relationship between a Japanese industry and US interest rates, and not Japanese rates. It’s also worth pointing out that the Japanese economy has had several false starts over more than a decade now, and so might not show any correlation with US/global monetary changes. That doesn’t explain why the 3-month rate would be significant, and the 10-year not, however.

  23. Kate, although I’m not entirely clear on what you’re doing, it doesn’t strike me as surprising that stock-market traders should be comparing asset returns to short rates, given that their holding periods are also short.

    Richard, I don’t have a clear answer, but as I will say in tomorrow’s Financial Review, I favour a combination of Asian central bank buying and a reduction in supply as the US Treasury shortens the maturity of the debt it issues.

  24. Fyodor,

    The time period I’m looking at is Jan. 1990- Dec. 2004 and I regressed proportional changes in my variables, i.e. I downloaded weekly prices of the metals and machinery sectors, and also respective bond yields from Bloomberg and calculated percentage changes of these.

    The reason I chose US rates is because the Japanese economy is heavily dependant on the US both as a source of raw materials for industry (particularly oil) and also as an export market. In the case of the metals sector, Japan not only imports a high percentage of the raw material from the US but also exports much of the finished product there. Therefore I would have expected it to exhibit sensitivity to changes in US rates which was the case, however I am still unsure as to why it was more so towards short term US rates.

    I have read that the japanese economy has experienced a number of “false starts ” as you described them over the last 10 years, and also that their fragile attempted recovery has been fuelled mostly by exports as opposed to domestic demand which is also why I choose to look at the relationship between the sectors and US rates. Also given the current macroeconomic climate in the US the issue seemed quite topical.


  25. Kate,

    That’s helpful. I think the USA is more important to Japanese metals and machinery companies as a market for their product, not so much as a source of raw materials. Metals, coking coal (for steel manufacture) and fossil fuels (coal and oil) are more likely to be imported from other countries (e.g. Australia, Brazil, the Middle-East) than the USA.

    The reason why I mentioned the number of “false starts” for the Japanese economy is that it has also had an effect on equity prices. That is, the performance of the Japanese metals and machinery index might have been flat, despite solid export performance, like the rest of the Japanese equity market over the 1990s. If the index had been flat over a period of considerable US interest rate volatility, this would dampen the apparent relationship between the variables.

    I can’t explain the greater significance of the 3-month yield over the ten-year rate, but my guess is that the short-term rate has shown greater sensitivity to economic growth than the long-term rate. If the share price performance of Japanese machinery companies is correlated with US economic growth, then this suggests that a proxy for US economic growth will show up as a significant variable. Remember that the US 3-month to a large extent anticipates changes in the US Federal Reserve funds rate (i.e. the official US interest rate), which is highly correlated with US economic growth.

Comments are closed.