Fundamental Analysis: Balance of Payments Analysis, Market Correlations, Sentiment Analysis

Balance of Payments Analysis

As I explained in Chapter 3, the impact of payments imbalances on exchange rates is anything but straightforward. In theory, a perennial trade deficit (when imports exceed exports) should cause currency depreciation so that long-term equilibrium can be restored. You can see from Figure 5-11, for example, that the release of UK trade figures in November 2011 touched off an immediate 150 PIP decline in the pound and seemed to catalyze an even bigger correction in the weeks that followed.

Figure 5-11. Impact of trade deficit on British pound

That’s because the figures showed a growing trade deficit and a goods deficit that was at the highest level since 1998. In other words, the markets concluded that at its current level, the pound was hindering both the recovery of the export sector (still languishing after the 2008 economic downturn) and the broader economy. To add fuel to the fire, the trade figures were even worse than the most pessimistic forecasts.

Naturally, the opposite should be true for any currency that boasts a trade surplus. South Korea and the rest of the so-called Asian Tigers, for example, have long experienced currency appreciation as a (undesirable) by product of their perennial trade surpluses.

Most countries release trade data broken down into goods and services, imports and exports, and on a monthly basis. Current account data, meanwhile, is typically released on a quarterly basis. In addition, there are an inexhaustible number of data streams for cross-border capital flows (classified by region, financial security, type of transaction, and so forth) which make excellent fodder for fundamental analysis!

In practice, it is tremendously difficult to derive the equilibrium exchange rate from a trade imbalance alone because the exchange rate is necessarily already at equilibrium. 

Recall that the difference between exports and imports should approximate the difference between savings and investment as well as the net change in ownership of domestic assets. This is basically another way of saying that trade deficits must be offset by net investment inflows, and vice versa. From this perspective, the relevant question is not whether net exporters are willing to subsidize net importers, but whether investors from net exporting countries are able and willing to deploy this difference into investment opportunities in the net importing economy at current exchange rates. In the words of one economic columnist, “When it comes to the U.S. trade gap, how many refrigerators the U.S. sells overseas is far less important than how many dollars the rest of the world wants.”

Figure 5-12, for example, shows that foreign entities purchase trillions of dollars of US securities (on a gross basis) every quarter, which provides a strong counterbalance to the trade deficit.

Figure 5-12. Gross quarterly purchases of US financial securities by foreign residents (Source: Bureau of Economic Analysis)

As an aside, current account deficits sometimes lead to bubbles in net importing countries, while other times it leads to currency depreciation. In the case of the US dollar, the perennial US current account deficit has caused both outcomes.

Another problem with analyzing the impact of trade data on exchange rates is that, by definition, they describe the past. As trade and investment flows reflect fundamental (non-speculative) shifts in demand for particular currencies, these shifts must necessarily have already happened prior to the release of the data. In other words, the fact that the United Kingdom experienced a large trade deficit last month doesn’t necessarily offer any clues into what will happen in the future.

Fortunately, this is almost irrelevant as far as you and I are concerned. Since forex trading is dominated by speculators, the fact that the markets adjust to trade and investment flows only after they have taken place (when they are reported) is not really a problem. In this way, these movements of money can be said to affect exchange rates twice: first, when they actually take place and, secondly, when the markets become aware of them. The first adjustment takes place gradually and often imperceptibly, while the second happens in large thuds.

Moreover, trade and investment flows typically follow easily identifiable trends. While deficits may spike from time to time, they typically move up, down, or sideways, and they do so for sustained periods of time. A structural shift from surplus to deficit (or vice versa) might signify that a currency is undervalued (overvalued), especially if it isn’t offset by a comparable shift in investment flows. In 2011, the Japanese yen rose to a record high against the US dollar, and its trade surplus steadily narrowed. The budding fundamental analysts out there might want to look for further clues that the yen might soon follow the dollar downward. (See Figure 5-13.)

Figure 5-13. (Potential) impact of changing trade patterns on USD/JPY (Source: Japan Ministry of Finance, Bank of Japan)

Sentiment Analysis

The theme of risk aversion was thrust to the forefront of investor consciousness in 2008 with the collapse of Bear Stearns and Lehman Brothers. As the banking crisis morphed first into a credit crisis and then into a full-blown financial crisis, the markets became transfixed by credit risk and moved to shed any assets for which there was even a slight possibility of default. This phenomenon manifested itself with especial intensity in the forex markets, where investors fled all emerging and peripheral currencies and migrated en masse into the US dollar. (Note Figure 5-14.) That’s because the United States is perceived as one of the safest places in the world to invest, due to the size of the US economy and depth of US capital markets. With this, the notion of the safe haven currency—one that is seen as preferable during times of crisis—was born.

Figure 5-14. USD, CHF, JPY: 2008 financial crisis and aftermath

Even after the markets recovered in 2009 and 2010, risk aversion was never far away. The sudden ignition of the European sovereign debt crisis, the earthquake in Japan, the downgrading of the US sovereign credit rating, and more all fueled investors’ fears. With each negative development, the markets responded in kind. The Japanese yen and the Swiss franc, which had also developed reputations as safe haven currencies, both rose to record highs. Simply, their capital markets are not deep enough to absorb sudden influxes of money without putting strong downward pressure on their exchange rates. The US dollar held up well, too. Throughout this period, there were plenty of developments that generated optimism. Each caused the flight of capital into safe havens to reverse. This manifestation of bipolar disorder was dubbed by one commentator as “risk-on, risk-off.”

In fact, risk aversion has long been a driver of currency markets, though previously it was risk appetite that hogged the spotlight. During most of the 2000s, record-high risk tolerance (some would call it complacency) led droves of investors into growth currencies. They came in search of higher yields, and the potential for currency appreciation was merely an added bonus. In 2011, the carry trade began to make a comeback, as investors were lulled back into a sense of security by rising interest rate differentials and economic recovery.

point of fundamental analysis, there are a few good quantitative indicators that are useful for forecasting risk. The first is simply forex volatility, which measures the extent of fluctuations in exchange rates and is used interchangeably with risk. Most forex portals offer volatility data on specific forex pairs, in absolute and percentage terms. In Figure 5-15, you can see how volatility in the EUR/USD has ebbed and flowed over time.

Figure 5-15. Average daily fluctuation in EUR/USD, number of PIPs (Source: Forexticket.co.uk)

The JP Morgan G7 Currency Volatility Index (shown in Figure 5-16) meanwhile offers the most comprehensive snapshot of overall forex volatility. A spike in volatility will typically precede a spike in risk aversion. That’s because an increase in volatility implies more variable (and therefore less dependable) returns.

Figure 5-16. JP Morgan G7 Currency Volatility Index, 2007–2011 (Source: Bloomberg L.P.)

Figure 5-16. JP Morgan G7 Currency Volatility Index, 2007–2011 (Source: Bloomberg L.P.)

Implied volatility in options contracts represents market expectations of volatility going forward. One can look at specific contracts in order to determine how volatility expectations differ across different currency pairs, time periods, and so on. With the Black-Scholes options pricing model, it’s possible to plug in all of the known variables and deduce the volatility that is implied by the price of the option. Most integrated quote/trading platforms (or a subscription to OptionVue) can perform this calculation automatically based on the other known parameters and display the implied volatility for any security/currency that has a corresponding option. Additionally, the New York Federal Reserve Bank regularly publishes data on implied volatility for major currencies, as shown in Figure 5-17. For example, the implied volatility of the JPY/USD in February 2012 was 10.1% (annualized) on a weekly basis and 13.5% on an annualized basis. When statistical theory is applied to these numbers, they imply a 68% chance that one year from now, the USD/JPY will be within 13.5% of the current exchange rate, and there is a 95% chance that it will fall within 27% (2 times implied volatility).

Figure 5-17. Implied volatility (%) for major currencies in February 2012 (Source: The New York Fed)

As for measuring the markets’ appetite for risk, the best proxy is probably the S&P 500 Index. When risk appetite is high, equities tend to outperform bonds, and growth currencies tend to outperform the majors. Similarly, the shift of capital from bonds into stocks (which is measured and released periodically by the mutual fund industry) also reflects an increased risk appetite. In fact, 2010-2011 witnessed a strong inverse correlation between the EUR/USD and the S&P 500. Each new revelation that the European debt crisis was deepening led investors to sell both US stocks and the euro. (See Figure 5-18.) As this correlation tightened, investors actually began to see the S&P 500 as a proxy for risk. As a result, big moves in the S&P 500 often preceded—not mirrored—changes in the EUR/USD.

Figure 5-18. Correlation between the EUR/USD and S&P 500

Market Correlations

Speaking of correlations, the currency markets are full of them. There are correlations between currencies and commodities, between currencies and financial securities, and between multiple currencies. Correlation can be discerned both through visual comparison and quantitative analysis. Sometimes, it’s enough to look at a chart of two different indicators (as with Figure 5-18) and immediately determine whether there is a relationship. Other times, it’s helpful to know the exact correlation coefficient, which, for statistics junkies, is the covariance of two data streams divided by the product of their standard deviations. A coefficient of 1 (or 100%) implies a perfect direct correlation, and a coefficient of -1 (-100%) implies a perfect inverse correlation. If the coefficient measures 0, then there isn’t any relationship between the two variables. While there is certainly disagreement over what the threshold is for significance, most would agree that a figure over 80% demonstrates a reasonably strong correlation and over 90% demonstrates a very strong correlation.

Of course, correlation does not imply causation. Just because two variables track each other very closely doesn’t mean that one necessarily causes the other. As with the observable relationship between the S&P 500 and EUR/USD, it could merely be that an external factor (risk appetite, in this case) is driving both to behave identically rather than implying that fluctuations in one are actually causing fluctuations in the other. This is a very important distinction because only instances of causation are actionable. Correlations can help us to understand the markets but are not entirely useful for plotting strategy. On the other hand, if I can determine that a rising S&P 500 Index is directly causing a rising EUR/USD, then I can buy the EUR/USD when the S&P rises and sell when it falls.

With that in mind, let’s look at some specific examples of correlation. Commodity prices tend to manifest themselves in currency markets in several different respects. Commodity currencies may take their cues directly from commodity prices, especially during periods of economic expansion. Canada, for example, is dependent on the United States for energy exports while China’s demand for coal and iron ore drives the Australian economy. As a result, the Canadian Loonie and the Australian Aussie are buttressed during commodity price booms. In the past, the South African rand has exhibited a correlation with gold prices, and the New Zealand dollar has always benefited from rising agricultural prices.

The role of oil prices in forex is slightly more complex. While there are a handful of economies around the world (namely, members of OPEC) that are completely dependent on oil, most are plagued by political instability and their currencies are not actively traded in the forex markets. The main exceptions are Norway and Mexico, whose respective krone and peso are both tied closely to the price of oil.

Much has been written about the relationship between oil prices and the US dollar. For most of the modern financial era, there was very little correlation between the two, probably because the price of oil didn’t fluctuate much. That changed around 2003 when oil prices began a 5-year, 400% climb, and the downtrend of the US dollar simultaneously accelerated (Figure 5-19). It was originally hypothesized that the latter caused the former. Since oil was priced in USD and the dollar was depreciating, oil producers had to raise their prices in order to offset the foreign exchange losses. Economists later determined that it was actually the other way around.

Figure 5-19. Rolling 6-month correlation between oil prices and trade-weighted USD (Source: U.S. Energy Information Administration, Federal Reserve Bank)

High oil prices negatively impact economic growth in the United States. In addition, the Fed’s core inflation index excludes food and energy prices, which means that rising oil prices will not likely be followed by higher interest rates, another negative in the short term for the dollar. Finally, trade between the United States and OPEC is largely one way, unlike trade between OPEC and the rest of the world, which means that US oil imports are not offset by increased exports. The upshot is that in the short term, significant changes (increases) in the price of oil can explain 50% (based on a correlation of -0.5) of subsequent changes (decreases) in the dollar, as seen in Figure 5-19.

Correlations between two or more currencies are the most interesting and the strongest quantitative relationships in forex. As I explained in Chapter 2, while there are dozens of liquid currencies, there is simply too much information for all of them to trade independently of one another. As a result, most currencies (especially those outside of the majors) tend to fluctuate relative to the US dollar. Emerging market currencies, in particular, behave similarly, especially during times when risk appetite is very strong or very weak. As can be seen in Figure 5-20, emerging market currencies rose and fell in lockstep for the first half of 2010. In the second half, risk appetite strengthened and growth/inflation differentials began to diverge, as did emerging market currencies.

Figure 5-20. Correlations in emerging market currencies break down

As far as correlations between individual currencies go, they tend to fluctuate in proportion to market conditions. As a general rule, however, currency correlations have been getting stronger over time. In Figure 5-21 it can clearly be seen that the correlation between the Canadian and Australian dollars has been relatively strong for most of the last decade and is currently nearing 100%!

Figure 5-21. Strengthening correlation between the AUD/USD and CAD/USD

There are a few theories for why this is the case, but the consensus is that currency markets are converging with other financial markets. Despite tremendous volume, forex trading was previously relegated to a quiet corner of finance. As more sophisticated traders expand into forex, they are bringing pre-existing trading mindsets with them. Many hedge funds, in particular, are connecting all of their trading operations under the umbrella of one broad strategy. The result is that algorithms may buy and sell the Canadian and Australian dollars together when commodity prices are rising, causing the correlation between these two currencies to become self-fulfillingly strong. Finally, the rising popularity of index funds (which consist of a basket of securities designed to represent a particular segment of the markets) has caused assets that were already correlated to become even more so.

There are a handful of online forex portals that publish real-time correlation data for the major currencies over different intervals of time. This information can be used toward a handful of strategic ends, which will be covered in Chapter 7. For now, consider that they serve as useful gauges for the strength of various fundamental indicators. For example, let’s say that one has a theory that rising interest rates are causing the Loonie to rise against the US dollar. Based on the matrix of correlations displayed in Figure 5¬22, however, it seems that the USD/CAD is quite strongly correlated (greater than 80%) with most of the other major currency pairs. This tells us, then, that the apparent rise in the CAD is better interpreted as a decline in the USD, and that one should probably look to US factors to explain the performance of the USD/CAD.

Figure 5-22. Weekly correlation between Canadian dollar and other major currencies (Source: Forexticket.co.uk)