Wednesday, August 20, 2008
Macro Man has finally breached the critical "feeling 80% of his best" threshold and is at long last back in the office for the first time since the first day of the month. Much has obviously changed since he last sat at his desk, not least the look of this very site. Feedback on the new template is welcomed; in addition to a slightly more polished look, it also includes a couple of new features such as the ability to search previous posts, which may prove useful.
In terms of markets, the most obvious change over the past three weeks has been the stunning rally in the US dollar. While Macro Man has previously noted that his anticipated three themes for August were in some ways correct, he will readily concede that he did not foresee the dollar's reversal of fortune.
Now that it's occurred, of course, the two obvious questions are 1) Why has it happened, and 2) What happens next? Macro Man intends to sketch out his initial thoughts below.
WHY HAS THE DOLLAR RALLIED?
While an exclusive focus on post-hoc explanation can be misplaced, Macro Man is a firm believer in the concept that you have to understand what you have seen before you can forecast what you're going to see. In that vein, Macro Man understands the dollar's rally as a confluence of five different factors:
a) Re-coupling. Over the past few months, Macro Man has been operating under an extremely useful analytical framework when it comes to analyzing currency markets. Put most simply, the market rewards those countries where the authorities are focused on fighting inflation, and punishes those currencies where the authorities are more focused on growth. Exhibit A in this regard is of course the US dollar, where Macro Man turned structurally and aggressively bearish nearly a year ago immediately after the Fed started cutting interest rates.
So what's changed, then? A couple of things. Of the popular major-currency alternatives to the dollar, some, such as New Zealand, have already begun to cut rates. Others, such as the UK, have strongly hinted that the prior easing cycle may start up once again. The RBA in Australia has surprised markets by also strongly hinting at imminent rate cuts. And while the ECB has not offered much hope for near-term rate cuts, by the same token they've at least toned down the rhetoric on the prospect for further hikes.
Moreover, the economic cycle in Europe is clearly weakening, and it seems clear that nominal GDP growth in Europe will soon dip below the policy rate for the first time in more than a decade. While the ECB can talk as tough as it likes, it seems increasingly obvious that at some point they will have to capitulate.
This de-rating of the alternatives to the dollar has, it seems to Macro Man, played no small part in bolstering the dollar's fortunes.
b) Commodity feedback loop. For a number of months now, Macro Man has held the unorthodox view that ECB tightening was exacerbating global inflationary pressures by applying pressure on the dollar, thus boosting commodity prices via the invoice currency effect.
Whether this view is correct or not is to some extent irrelevant; what does seem clear is that commodity prices and the dollar had become entangled into the same trade. Thus, when oil started to move lower thanks to lower global demand, an inventory boost, technical selling, etc., this provided a boost to the dollar. Oil traders, seeing the dollar rally, then dumped more oil (and gold traders dumped gold), encouraging a further dollar rally.
Structural commodity longs have, according to the market intelligence that Macro Man has received, bought USD in the currency market as a hedge. What there can be no dispute of, however, is that currency people have pointed at commodities as a reason to buy dollars, and commodity people have pointed at the dollar as a reason to sell oil/gold, etc. Certainly Macro Man has witnessed an amusing repartee on Bloomberg chat between FX and commodity salespeople from the same bank, with each pointing at price action in the other's market as a rationale for movement in their own bailiwick.
As the vicious commodity/dollar cycle turned virtuous, the dollar benefited.
c) Hedging by long-term structural shorts. For the past few years, US investors in foreign markets have had the ultimate free lunch. Not only has topline nominal GDP growth in the rest of the world exceeded that of the US (thereby implying higher local currency asset market returns), but the dollar has weakened substantially as well. Whichever way you looked, these investments have been winners.
However, with growth in the rest of the world slowing and formerly bulletproof markets showing signs of mortality (Bovespa, anyone?), two things have happened. Assets in dedicated non-US funds have started to cash in. This is very imperfectly proxied in the graph below, which charts shares outstanding in the MSCI Emerging Markets ETF available to US investors (Macro Man held it in his personal account until late last year.)
As you can see, there were quite substantial redemptions during the period of equity market weakness from May-July. Although inflows recovered along with stock prices, they never matched previous highs, and recently they have turned south again.
Perhaps more importantly, however, long-term US holders of foreign assets, particularly assets in the rest of the developed world, have realized that they now have an opportunity to lock in extremely attractive exchange rates. While this will not mitigate against declines in the local currency values of foreign assets, it at least provides protection against being shot with both barrels (e.g., a fall in the local currency value of an asset AND a decline in the value of the local currency against the base currency.)
Shortly before he went on his holidays, Macro Man had begun to hear rumblings of "long term real money selling euros that they've held for five years." Subsequently he has heard from a number of counterparties that this flow continued and indeed accelerated during his absence. This type of flow provided support for the dollar at overvalued levels in 2001 and early 2002; with the dollar now undervalued against many currencies, small wonder that it is helping spur a recovery.
d) CTAs awaken. Benjamin Franklin famously said that nothing is certain but death and taxes. Allow Macro Man to add a third item to that list. If a financial asset, particularly a currency, moves far enough, fast enough, trend following models will kick in and exacerbate the trend. The chart below shows the EUR/USD exchange rate with Bollinger bands, which are basically standard deviation intervals around a moving average.
Of particular interest here is the middle section of the chart, which shows the band width. Observe how it spent a number of months at a relatively low level. In this sort of environment, many trend following models switch off, as a narrow band width suggests mean-reverting rather than trending markets. However, the sharp rise in the band width over the last few weeks will have re-ignited these models, who would then have sold EUR/USD with gusto, thus helping the dollar to strengthen.
Indeed, Macro Man has had this experience with his own trend following model, which doesn't use Bollinger bands at all.
e) Addition by subtraction: FX reserve managers find something better to do. Longtime readers will know of Macro Man's firm belief that reserve accumulation shenanigans by a number of emerging market countries (China, India, Russia, the Middle East) has exerted a substantial upward influence on the EUR/USD exchange rate.
Recently, however, the piss-taking activities of many of these countries has waned and even reversed as they have reaped the inflationary harvest sown by their serial local currency selling. Indeed, a number of countries have seen their currencies weaken to the extent that they are even selling reserves into the market. All else being equal, this would imply a need to buy USD against other developed market currencies to maintain portfolio benchmarks.
As Brad Setser would tell you, not else is necessarily equal; a decline in the EUR/USD rate reduces the need to sell EUR/USD via the valuation effect. Still, it is fair to say that the CBs have been refreshingly absent from the G10 foreign exchange market recently; call it addition by subtraction.
SO WHERE TO FROM HERE?
Ah, the million (billion? trillion?) dollar question. At this juncture the list of what Macro Man does not know is substantially larger than the list of what he does, which naturally reduces his confidence in any of the conclusions that he may reach. A useful starting point may be to consider how durable some of the above-mentioned factors may be. Very briefly:
a) Re-coupling. From Macro Man's perch, this set looks to continue, at least within the developed market. It's therefore likely to remain dollar-positive moving forwards.
b) Commodity feedback loop. This could fall out either way. If China cranks up the factories after the Olympics and/or hurricane season is a bad one, oil could be back above $130/bbl before you can say "Sell my SUV." By the same token, if the weather is a non-event and Chinese industrial activity continues to decelerate, we could easily threaten or perhaps breach $100. Confidence is low on this one, but for choice Macro Man would suggest that from here commodities are more likely to be a dollar negative than a positive.
c) Hedging/repatriation. In many ways this is a similar phenomenon to item a), and it therefore seems reasonable to expect it to continue to support the dollar.
d) CTAs. A bridge made of balsa wood provides more enduring support than these chaps.
e) CBs. Macro Man's best guess: they return to market (after all, most of these guys are still running tasty current account surpluses), but with a lower profile moving forward than they've had in the past. The acid test here will be Q4 of this year; for most of this decade, CBs have been very vigorous dollar sellers against G10 in the fourth quarter of the year.
Adding it all up, Macro Man arrives at the conclusion that the buck is likely to be more stable in the future than it has been in the past. He would be very, very surprised if we revisit the dollar's lows against the rest of G10 during this cycle.
A useful historical analogue is USD/JPY in the mid 1990's, which was a seminal event in framing Macro Man's understanding of long-term currency cycles. As readers may or may not recall, the early 1990's saw USD/JPY drop steadily from 160 to 100 by late 1994, and then collapse to 80 in the spring of 1995. After jiggling around the low 80's for a few months, in the summer of 1995 it then surged back to 100 as quickly as it had fallen, which history eventually showed was the initial phase of a new dollar bull trend.
The reversal of fortune in EUR/USD struck Macro Man as exhibiting similar tendencies, but even he was amazed when he overlaid the last few years' price action in EUR/USD with that of USD/JPY in the early-mid 1990's:
The key price action here is obviously the "hump" in the upper-right hand corner, which is remarkably similar in both shape and duration for both currency pairs.
While historical analogue overlays can be over-interpreted, in this case Macro Man believes there is a reason to believe the parallel, as in both cases the dollar has sharply recovered from badly over-sold (and under-valued) levels, thereby dispelling the notion that dollar weakness is a one-way, ever-profitable bet.
Now, the fundamental backdrop now versus the mid-1990's is completely different, so it's unreasonable to expect the analogue to hold tick-for-tick. But the follow through in 1995-96 for USD/JPY- a year of sideways, range-trading, back-and-fill price action- seems to be a reasonable forecast for the next few months and quarters for the dollar.
Whether the buck can eventually rally against the euro as it did against the yen in 1997-98 is another question altogether, and Macro Man suspects that the answer is "no."
The degree to which the dollar can "buck up" from here will, in all likelihood, ultimately depend on the extent to which the US economy and financial system can recover from the current crisis; issues which seem unlikely to see a quick resolution.
Still, the dollar moving from a one-way bet to a two-way bet is a significant change, and one that Macro Man intends to heed in his investment strategy. This month Macro Man has already taken the biggest dollar long bet that he's had in two or three years, and while he can certainly contemplate going short again, perhaps soon, from his perspective the era of the "all-in" bet from the short side is over.