FX Crossroads: The Puzzle of the Missing NOK Rally

Summary and conclusions

• The NOK's recent performance has failed to meet expectations - ours included. Despite a surge in oil prices and a recent upswing in the domestic stock market the NOK has remained under pressure. The correlation with other high-yielding currencies has been high and a substantial risk premium has been attached to NOK. Looking forward, a return in risk appetite will most likely support NOK, which is also backed by sound domestic economic fundamentals.
• In December 2007 we identified a slowdown in the global economy. On the back of that, we anticipated JPY, EUR and CHF to deliver positive returns and AUD, GBP and CAD to deliver negative returns. Buying this basket would to date have delivered an annualised return of 14%, with a Sharpe ratio of 1.13. We expect a continuation of the slowdown in the months to come and the signals of our business cycle analysis to be unchanged.
• Volatility has fallen and carry has performed well since the middle of March. We expect volatility to begin trending higher again in May, bringing carry outperformance to an end.
• Our EUR/USD target of 1.60 has been reached. We will publish new FX forecasts on Monday, 5 May.
• FX Crossroads is published every second Wednesday. Next publication date is 14 May 2008.

Link between oil and NOK less clear recently - 6-month change in effective oil price and EUR/NOK

Our business cycle analysis has proven robust
out of sample

NOK: The Puzzle of the Missing Rally

NOK: Wolf in sheep's clothing?

The NOK's recent underperformance has failed to meet expectations - ours included. In this article we examine if the oil price is still the dominant factor for NOK and if the currency's relationships to other asset classes is any stronger. We conclude that the surge in oil prices has had only a minor impact on NOK while the level of risk aversion has become a more important driver. We also examine whether NOK - with a 3-month interest rate well above 6% - has started to behave like a high-yielding currency and should therefore be treated as one. Here our conclusion is mixed: we identify both non-carry and carry characteristics in NOK.

Is oil still driving NOK?

The oil price affects NOK in a number of ways, and this remains true despite the implementation of various strategies to reduce its sensitivity to swings in the oil price. Norges Bank (NB) tries to offset these flows by selling NOK against foreign currency, but a seasonal pattern persists (see eg, NOK: The influence of oil taxes, 25 March 2008). In some periods this selling is, however, more than matched by NOK purchases by foreign banks, which generates NOK demand and a potential appreciation. The correlation between NOK net purchases and EUR/NOK has historically been quite strong (R2: 0.56) but has recently been weaker; currently it implies a stronger NOK.

Another example of the strong historical link between the oil price and NOK can be seen in the 6-month change in the effective oil price vs. the corresponding change in the EUR/NOK exchange rate (R2: 0.49). A simple rule of thumb (since 2005) implies that a 10% increase in the effective oil price yields a 1% fall in EUR/NOK.

However, this clear link has also suffered recently. The effective oil price is more than 30% higher than it was six months ago (which under 'normal' circumstances would imply a 3% strengthening of NOK vs the EUR) but NOK has in the same period lost 3%. Accordingly, NOK must recently have been driven by other factors than the oil price.

While the effective oil price has doubled since January 2007, the correlation between NOK and oil has deteriorated. In January 2007, the correlation was around 30% and has since slumped to around 20%. In the same period, NOK has moved more in line with other, previously uncorrelated, asset classes.

Equity prices have been a far more important factor than previously when explaining NOK movements in H2 2007 and the first months of 2008. The 100-day rolling correlation between NOK and the domestic stock market reached 50% in November 2007. In a broader context, risk aversion seemed to dominate, keeping NOK under pressure despite sound domestic economic fundamentals.

The strong correlation between currencies and equities is unlikely to persist in the longer run; a 'normalisation' of relationships in currency markets will probably occur, i.e. a weaker link to equity prices and a stronger link to relative interest rates. This indicates a shift away from risk aversion towards more risk-seeking behaviour. Such a 'normalisation' has in fact already begun, but without NOK benefitting significantly. In April, the OBX has outperformed other European equity indices without a rally in the NOK. One could therefore wonder if NOK has gained new characteristics.

NOK: mostly low- or mostly high-yielding?

Since February, NOK has been the third-highest yielding currency in the G10 currency universe and has therefore been established as a carry-target currency. The G10 carry basket (defined as the three high-yielding G10 currencies vs. the three low-yielding ones) has had a chequered career in 2008 but is broadly unchanged over the year. The Norwegian currency market is, however, quite small and carry flows can potentially impact NOK substantially.

Volatility in FX markets has broadly risen in recent months, but NOK volatility (vs. EUR) has increased less than that of other currencies, e.g., USD, GBP and AUD, and NOK uncertainty is now perceived as lower. One exception to this is the SEK; NOK is now regarded as more sensitive to investor sentiment. The message from volatility ratios is that generally NOK is not perceived as a highly volatile currency.

An opposite example can be found in actual FX correlations. From these we observe that most NOK correlations have become more pronounced which also was the situation around three years ago. That the SEK/NOK correlation is strong may not be a surprise, but the strong negative relationship with the CHF (and the USD) and the very strong positive relationship with the AUD are prominent. These indicate that NOK has less in common than it previously did with what are widely regarded as safe-haven currencies, and has more in common with riskier currencies.

Whether NOK has attracted more attention from speculative investors due to a higher interest pick-up remains an open question. This is because we do not have any data on speculative NOK positions, contrary to e.g., AUD positions. It is, however, highly likely that investors largely prefer to build up long NOK positions when risk appetite increases and reduce positions when risk aversion strengthens.

Currently, we cannot detect whether NOK has mostly low-yielding or mostly high-yielding characteristics. On the one hand the Norwegian yield spread vs. G10 is positive and increasing and NOK has started to move more in line with riskier currencies. On the other hand, NOK volatility is perceived to be relatively low going forward, which isn't exactly a carry characteristic. To get good evidence, we still need to see how NOK behaves in a risk-seeking environment.

Summing up, we believe that there has been a substantial risk premium attached to the NOK recently and that investors have priced risk more aggressively than previously. Our anticipation is that relationships with other asset classes will normalise and that oil and interest rates will favour NOK going forward. A return in risk appetite will most likely support NOK which is also backed by sound domestic economic fundamentals.

G10: Business Cycle Update

Strong model performance

Linking FX performance to the business cycle

In December last year we presented an analysis of the relationship between currency movements and the global business cycle; see FX Strategy: G10: Business cycles and FX performance, 10 December 2007. In this article we take a look at how the model has performed since December and review the current signals of the model.

Overall we can conclude that the model has performed well and that the main results so far have proven robust out of sample. On 7 December the OECD leading indicator covering October was released and shifted from signalling a downturn to signalling a slowdown of the global economy. In such a scenario the model identifies JPY, EUR and CHF as having historically delivered positive returns and AUD, GBP and CAD as having delivered negative returns. Buying this basket would to date have delivered an annualised return of 14%, with a Sharpe ratio of 1.13. We expect the CLI to continue to signal a slowdown in the months to come and the model signal is thus unchanged.

A recap of the analysis

In accordance with OECD research we have defined four states of the global business cycle based on the OECD composite leading indicator (CLI). These are shown in figure 1 below.

We have then looked at the historical average return on each of the G10 nominal effective exchange rates in the different states of the business cycle, and identified those currencies which are systematically related to the business cycle. A summary of the results are shown in the table below, where winners (losers) are those currencies which deliver a significant positive (negative) return. As always, we remind our readers that any purely empirical relation should be considered with caution, since it is not directly explained by theory and thus could be spurious.

Very strong performance

On 7 December last year the CLI covering October was released. The October reading showed a fall to 99.3 from 99.7 in September, and the economy thereby moved from a state of downturn to a slow-down. In a slowdown the model identifies a significant positive return for JPY, EUR and CHF, while identifying a significant negative return for AUD, GBP and CAD. Figure 2 shows the change in the nominal effective exchange rates following October, when the economy moved into a slowdown.

During this period we can observe that the overall exchange rate movements have been consistent with the model findings. The JPY, EUR, and CHF have all appreciated, while the CAD and especially the GBP have weakened. The only fallout of the analysis has been the AUD, which is approximately unchanged over the period.

One way of applying the model is to buy a basket of currencies defined by the significant signals of the model. Below we have shown the index return of this strategy, when it is assumed that the currency basket is bought on the day of the model signal. As can be seen from the graph, this would not have been a bad strategy. An investor who had applied this investment rule would thus have earned an annualised return of 14% to date. Even though the return would have been somewhat volatile, this has been compensated for by a Sharpe ratio of 1.13.

The signal going forward

Since last fall the CLI has been falling steadily. And although the latest observation covering February showed a small increase on a month-to-month basis, the 3-month change is still negative.

The CLI is thus still signalling a slowdown and the model is therefore also still implying that JPY, EUR and CHF should outperform, while AUD, GBP and CAD should underperform

We are not yet looking for a shift in the CLI

While the February CLI reading edged slightly higher, we do not expect this to mark a trough of this business cycle. We continue to believe that the global slowdown has further to run and that it is premature to call a turnaround in just six months (the CLI is leading industrial production by six months). However, should the CLI shift to signalling a recovery, then the model will be generating a very different signal. The model would then imply that AUD, NZD and CAD should outperform, while CHF and EUR should underperform.

Conclusion

The link from business cycles to exchange rate movements goes through as different channels as monetary policy, production growth, productivity changes, flow of funds, and relative external balances. It is therefore naïve to believe that this link can be captured by any simple model. However, despite the simplicity of this model, the signal appears surprisingly strong during periods of negative output gaps, and around times of a cyclical shift in the business cycle.

The current signal of a continued economic slow-down is overall consistent with our general views. We continue to forecast a stronger JPY and CHF, while we believe that both the GBP and CAD weakness has further to run. We do believe, however, that a shift in the European business cycle is beginning to unfold, which will leave the EUR under pressure. We are furthermore not bearish on AUD in the very short-term, as we expect relative economic strength, a rising terms of trade, and favourable relative interest rates to lend further support. We are, nonetheless, still forecasting a lower AUD by year-end.

G10: Implied Vol Set to Rise in May

Vol set to rise, carry to underperform

Implied FX vol has risen in four distinct waves since August 2007. Each wave has seen a correction lower while keeping the trend towards higher vol intact (chart 1). Since mid-March USD/JPY implied 1m vol has fallen from 16.3% to 11.25%. The decline in volatility has coincided with a rise in risk seeking in financial markets overall (chart 2). We do not expect volatility to fall much from here and look for vol to trend higher in May. In general, we believe that volatility is likely to rise on a trend basis due to financial deleveraging and the ongoing global economic slowdown.

Consistent with the decline in volatility, G10 carry has performed well in the past month. A G10 carry index (long 3 high yielders vs 3 low yielders) has gained more than 5% since mid-March. In annualised terms the 1-month return is approximately 75% (chart 3). In recent history, such performance levels have coincided with a period of carry under-performance. It is always difficult to pinpoint the exact turning points, but the recent carry performance nonetheless raises a yellow flag for the coming period.

As the final chart below highlights, carry performance and implied volatility goes hand in hand, that is, a rise in vol will coincide with carry underperformance. This is not exactly surprising since higher volatility lowers the carry-to-risk.

If we are right in expecting a turn in volatility as well as in carry performance, JPY and CHF looks likely to gain on high yielders such as AUD, GBP and NZD. Strictly speaking, USD is a funding currency in G10 terms, but higher vol should see USD/JPY move lower.

Full Report in PDF

Danske Bank
http://www.danskebank.com/danskeresearch

Disclaimer

This publication has been prepared by Danske Markets for information purposes only. It is not an offer or solicitation of any offer to purchase or sell any financial instrument. Whilst reasonable care has been taken to ensure that its contents are not untrue or misleading, no representation is made as to its accuracy or completeness and no liability is accepted for any loss arising from reliance on it. Danske Bank, its affiliates or staff, may perform services for, solicit business from, hold long or short positions in, or otherwise be interested in the investments (including derivatives), of any issuer mentioned herein. Danske Markets' research analysts are not permitted to invest in securities under coverage in their research sector. This publication is not intended for private customers in the UK or any person in the US. Danske Markets is a division of Danske Bank A/S, which is regulated by FSA for the conduct of designated investment business in the UK and is a member of the London Stock Exchange. Copyright (©) Danske Bank A/S. All rights reserved. This publication is protected by copyright and may not be reproduced in whole or in part without permission.