New research suggests way to make steady profits from
the carry trade:
Crash and carry
When pundits worry about the distorting effects of cheap money on
asset prices, they invariably single out the carry trade as a cause for
concern. The term is often used loosely to describe any investment that
looks suspiciously profitable.
More specifically it refers to a particular sort of foreign-exchange
trading: that of borrowing cheaply in a 'funding' currency to exploit
high interest rates in a 'target' currency. The yen has long been a
favoured funding currency for the carry trade because of Japan's
permanently low interest rates.
As a result of the crisis and near-zero rates in America, the dollar
has become one, too.
If markets were truly efficient, carry trades ought not to be
profitable because the extra interest earned should be exactly offset by
a fall in the target currency. That is why high-interest currencies
trade at a discount to their current or 'spot' rate in forward markets.
If exchange rates today were the same as those in forward contracts,
there would be an opportunity for riskless profit.
Arbitrageurs could buy the high-interest currency today, lock in a
future sale at the same price and pocket the extra interest from holding
the currency until the forward contract is settled.
In practice, the forward market is a poor forecaster. Most of the
time exchange rates do not adjust to offset the extra yield being
targeted in carry trades.
So a simple strategy of buying high-yielding currencies against
low-yielding ones can be rewarding for those that pursue it.
The profits are volatile, however, and carry trades are prone to
infrequent but huge losses.
In late 2008 the yen rose by 60% in just two months against the
high-yielding Australian dollar, a popular target for carry traders.
That made it much more expensive to pay back yen-denominated debt.
If efficient-market theory cannot kill the carry trade, why don't
volatile returns, and the occasional massive loss, scare off investors?
A new paper by Oscar Jorda and Alan Taylor of the University of
California, Davis, may have the answer.
They find that a refined carry-trade strategy-one that incorporates a
measure of long-term value-produces more consistent profits and is less
prone to huge losses than one that targets the highest yield.
The authors first examine returns to a simple carry trade for a set
of ten rich-country currencies between 1986 and 2008. Buying the highest
yielder of any currency pair produced an average return of 26 basis
points (hundredths of a percentage point) per month.
That would be fine, except that the standard deviation of returns, a
gauge of how variable profits are, was almost 300 basis points.
The monthly Sharpe ratio that measures returns against risk was a
'truly awful' 0.1 (the higher the ratio, the better the risk-adjusted
performance).
Worse still, the distribution of monthly profits was negatively
skewed: big losses were more likely to occur than windfall gains.
No sane trader would follow a rule with such poor results. So the
authors put together a far richer model to help decide which side of a
currency trade to be on.
It included things that are most likely to influence short-term
movements in currencies, such as the change in the exchange rate over
the previous month, as well as the size of the interest-rate and
inflation gap between each currency. They found that all three factors
mattered.
Currencies that rose in one month tended to rise in the next month.
Those with the highest interest rates went up most, as did currencies
with high inflation (which drives expectations of further rate rises).
These impulses can drive exchange rates a long way from their fair or
'equilibrium' values. That creates the risk of a sudden reversion that
could wipe out earlier profits.
To guard against this, the authors added to their model a measure of
how far the exchange rate has shifted from its fair value.
They found that this alarm bell can sometimes turn a 'buy' signal
into a 'sell'.
This valuation check helped get rid of the negative skew associated
with the simplistic version of the carry trade.
But the authors thought the model could be improved still further.
One worry was that although it makes sense for traders to buy currencies
with fat yields, it may be dangerous past a certain point. After all, a
high interest rate can be a symptom of a currency in distress.
The authors judged that the link between profits and yields was
likely to be 'non-linear' (ie, its strength alters as the interest rate
of the target currency climbs) and changed their model to reflect this.
This non-linearity applies to currencies' values, too: the likelihood
of a crash escalates as a currency becomes ever dearer.
Sharp thinking
The fully evolved model performed well compared with its primitive
ancestor. Used in a portfolio of separate carry trades (to limit the
volatility that comes with making a single bet), it delivered average
monthly returns of 57 basis points, much better than the 26 basis-points
profit from the simple approach.
The Sharp ratio based on annual returns was a very healthy 1.27. And
returns based on the deluxe model had a positive skew: large windfalls
were more likely than big losses.
It appears that savvy investors can indeed make sustained profits
from the carry trade. The authors stress that their sophisticated
approach was scarcely better than the simple one at predicting the
direction of exchange rates. The crucial advantage is that where it made
mistakes, the stakes were small.
The deluxe model might tell a trader to turn his nose up at a trade
with apparently strong returns because of the risk of a currency crash.
The trade might well turn out to be profitable but the forgone profit
is a small price to pay for avoiding a potentially big loss.
It is a lesson that applies to other asset markets, including those
for shares, bonds and homes.
Trading momentum will often drive up asset values for long periods
and persuade buyers that high prices can be justified.
But investors ignore fair-value measures at their peril. The
Economist
|