The Risk of $ repatriation

This week might be for some of our readers a contreversial topic. We would welcome any feedback on this because we can imagine this is going to raise some eyebrows. At this very moment the theme of the day remains the risk aversion of investors supporting the $. Looking further down the road there still is this phenomenal and exponentially rising debt the US is carrying and which is funded by the rest of the world. This is a serious risk hanging over the markets and could lead to a collapse of the currency. There is a broad consensus to support this view.

However during the research we did over the last couple of months we came across a couple of influential writers who threw a different view on this. We did our number crunching and must admit there is a point to be made by saying that the risk of a total $ collapse might be not that high.

Earlier last week there were already some very encouraging comments from Chinese government officials arguing they keep on having faith in and commit themselves towards their further investments in US Treasuries. Usually there is a gap between rhetoric and facts however the recent auctions in US Treasuries confirm these statements and show that investors in general keep on buying US sovereign paper.

In order to finance the current account deficit foreign investors have been willing to buy a record number of US assets. A detailed overview for 2008 is not available yet but the overall number of foreign purchases of US portfolio assets is at $ 9.5 tln. Table 1 shows an overview of the different assets held by investors outside the US.

Table 1: Foreign Holdings of U.S. securities, by type of security


Source: United States Dpt. of the Treasury

If we compare this with the total value of US citizens holding foreign securities we come out at a number of roughly $ 5 trio. This is down from $ 7.21 tln at the end of 2007. Two important things do strike us when looking at these numbers. First of all the amount of US investments foreigners are holding is almost twice as high as the amount of foreign securities held by US citizens.

Secondly the drop of foreign assets held by US citizens versus the continuous rise of foreign purchases of US assets. The more than $ 2 tln drop of foreign assets held by US citizens explains the recent rally of the $ versus EUR and £ as American investors repatriated their money over the last year rapidly. Especially in times of crises this is a typical phenomenon. Chart 1 shows the net flows of US purchases of foreign securities, with a negative number indicating a net outflow of capital from the US vice versa a positive number showing a US sales of foreign securities (on a month-to-month basis).

Chart 1: Net in versus outflows for American Investors



Source: United States Dpt. of the Treasury

Since the late nineties we see more volatile spikes. 1998 was accompanied by the Asian crisis and the implosion of LTCM, followed by late 2000 the beginning of the dotcom bubble bursting and 9/11 causing some distress among investors as well. All these periods went hand in hand with a rally in the $ across the board.

As history shows, US investors have the tendency to shy away form foreign markets for several years, after being hit by a shock. This backs our view that for the time being the $ will remain very well supported.

The current crisis clearly underlines once again (as supported by other data) the seriousness of the situation.

The good news for the time being is that foreigners are still willing to buy US assets despite the credit crunch, Fed rates going to zero and a worrisome mounting level of US Government debt. The official comment from the Chinese Government is a clear example of this and very important towards other big institutional players who consider China the leading player at the table.

However, to what extent is the risk of a repatriation of foreign investors money realistic, and could the impact be significant? Assume that ten percent of the USD 9.5 tln held by foreigners would be repatriated during a year of extreme events and let’s assume the US current account deficit would stay around the same current levels of $ 600 bio.

The end game would be the US economy suffering a balance of payments deficit of appr. $ 1.55 tln. This in turn would require American investors to repatriate 30% of the current $ 5 tln held in foreign assets to prevent the $ from collapsing. Last year the brief fall of the $ versus the EUR from 1.25 towards 1.48, may have been a simple strategic adjustment from the central bank of China by $ 60-65 bio of bonds that were shifted in European sovereign bonds. We will come back to this later on whether this is a valid point.

The next question we need to ask is whether this risk is realistic? As Table 2 shows, foreigners hold over 55 % in US Treasuries. However compared to the total amount of outstanding long-term US Securities this is less than five percent. If we add this up to the US government agency securities we still come out at a number far below ten percent of the total US capital market.

So the argument whether the high dependency of the US Treasury debt from foreign institutional investors needs to be tempered. Given the flexibility of the US capital market there is enough room of reallocation of investments by different groups of investors among different asset classes. If foreigners decided to shift their holdings of treasuries other investors would be attracted by the reduction in prices of treasuries to make switches in the opposite direction. There might be some alteration in the relative prices of the different US assets, with a modest increase in the cost of financing the federal debt, but major disruptions would be highly unlikely. This is more from an interest rate perspective.

The impact on the FX rate of the $ towards other currencies will be less positive but should also not be overestimated. It is plausible that we could see a move of 20-30%, which is already huge, like we have seen in £ or like the EURUSD move from late August last year. However a sudden unwind by e.g. the Japanese ($ 578 bio) or Chinese authorities ($ 696 bio) of their total US Treasury reserves hand in hand with a total implosion of the $ is at this stage less probable. One should not forget that both countries, who hold the largest part of US government bonds, are very much in favour of avoiding market disruptions and (this is more unfortunate) favour to maintain undervalued exchange rates to bolster even further international competitiveness.

Then there is the liquidity argument of the Foreign Exchange markets. Based upon the data from the BIS the daily volumes of the FX market were $ 3.21 tln in April 2007 (this is a survey held every three years). The FX market is by far the most liquid market in the world.

Already intuitively you might see that it is not correct to draw conclusions on flow info that e.g. a $ 60 bio bond issue coming to maturity from the Bank of China would cause a movement from 1.25 to 1.48. There are a couple of reasons for that.

As far as we know there are no theoretical studies on the impact of volumes on the price action of a currency pair. Even empirical studies are very limited since there are not a lot of data available on foreign exchange traded volumes at high frequencies. As mentioned above the BIS offers surveys on this subject only every three years.

Literature on equities and futures though gives us an indication that volumes and spreads are positively correlated. So basically we can use the bid-offer spread as an indicator how a market can swallow certain flows. As a point of reference if you would be active on a corporate desk responsible for the execution of FX flows of companies and other institutional clients, a spot trader would quote you a bid offer spread of 50 – 70 points for an order of $1 bio. One can not extrapolate this linearly for an amount of $ 60 bio but we can say for sure that a $ 60 bio amount will not move the markets by 2,300 points. One has to take into account that in a $ 2.3 trio daily market by nature there will be similar interests at the other side which will absorb a substantial part of this.

Then there is the evidence that US investors repatriated appr. $ 2 tln during 2008 back to the US. This information by itself counters the argument that a $ 60 bio flow would move EURUSD from 1.25 to1.48. Let’s even assume this would be the case, and let us assume the repatriation would have started at the top of the market, i.e. 1.60. If that would have been the case a $ 2 tln flow should bring EURUSD much lower then 1.25 at the moment.

Furthermore, based upon the above information, if the Bank of China would decide to sell its total portfolio of US Treasuries for an amount of $ 695 bio, we believe the market would be mature enough to absorb this flow. This does not mean it would not have an impact, but it would certainly not create the total implosion of the $ as many claim. There could even be some follow through from other market participants who join the move. But let’s even assume this would cause a total outflow out of the US of $ 4 tln (which is substantial) it would probably only have a similar impact as the recent repatriation from US investors we have seen.

For a total implosion one would need a complete repatriation of foreign investors US Treasury holdings. The question would be however: where to? Europe with a similar rising debt and the PIGS countries on the brink of default? The UK with a similar imploding banking system? Japan with its deflationary vicious circle? Or China which does not want a stronger renminbi and far from a transparent and developed country?

Table 2: Table 3 Value of foreign-owned US long-term securities and share of total outstanding, by asset class



Source: United States Dpt. of the Treasury

(1) Fred Bergsten and Edwin Truman “Why Deficits Matter: The International Dimension.” Peterson Institute of International Economics, January 2007.