Monthly currency seasonality: Down and out?

Olivier Desbarres

Share on twitter
Share on whatsapp
Share on facebook
Share on linkedin
Share on email

Listen to our reports with a personalized podcasts through your Amazon Alexa or Apple devices audio translated into several languages

Report Contents

( 14 mins)

This report updates the monthly seasonal patterns of 31 major Nominal Effective Exchange Rates (NEERs) going back to January 2010, using over two million daily data points with trade-weights derived from the BIS (April 2019) and national central banks (see “Nominal Effective Exchange Rates: Monthly seasonal patterns”, 10 January 2019).

A number of factors can drive currency seasonality, including underlying seasonal patterns in balance of payment flows, the timing of public holidays and market liquidity.

However, significant changes in central bank or government policy, major domestic events (e.g. Brexit) and once-in-a-lifetime “black swan” events such as the Covid-19 pandemic can break down even well established seasonal currency patterns. In emerging Asia interventionist central banks have in part eroded monthly currency seasonality.

Unsurprisingly, typical monthly seasonal currency patterns broke down in 2020, with a few exceptions. In particular, when global risk sentiment collapsed in March and peaked in June only a handful of currencies moved broadly in line with their historical patterns.

Historical seasonal patterns somewhat re-established themselves in Q3 (notably August) with domestic factors once again driving asset prices, including currencies. But the resurgence in Covid-19 cases and deaths in Q4 along with the re-imposition of national lockdowns once again contributed to a breakdown in historical seasonal patterns, albeit not as marked as in March or June.

The Brazilian Real and Colombian Peso NEERs recorded the strongest average rates of appreciation in the month of January in 2010-2019, while the Turkish Lira and Russian Rouble were the weakest. About 15 NEERs traded broadly in line with their historical patterns in January 2020.

However, the Covid-19 pandemic was still in its infancy and its impact on global economies and financial markets still modest. The risk is that many currencies, including the typically strong Swiss Franc, may in January 2021 continue to deviate materially from their 2010-2019 seasonal patterns in our view.

 

The case for looking at Nominal Effective Exchange Rate (NEER) seasonality

This report updates the monthly seasonal patterns of 31 major Nominal Effective Exchange Rates (NEERs) going back to January 2010 (i.e. 132 months), using over two million daily data points, with trade-weights derived from the April 2019 Bank of International Settlements (BIS) report and national central banks (see Nominal Effective Exchange Rates: Monthly seasonal patterns, 10th January 2019).

The NEER is a trade-weighted average of nominal bilateral rates between a country’s currency (e.g. the Dollar) and the currencies of this country’s main trading partners (in the case of the United States, namely the Chinese Renminbi, Euro, Mexican Peso, Canadian Dollar, Japanese Yen, Korean Won and Sterling). For this reason the NEER is also referred to as the Trade Weighted Index (TWI). Please see Figure 13 for further details about our calculations Methodology.

Most (if not all) studies of currency seasonality tend to focus on crosses versus the Dollar (e.g. EUR/USD, USD/JPY) as they are amongst the most actively traded crosses. Moreover, in many cases the Dollar has a large weight in currencies’ NEERs as the United States is a major trading partner for many economies (for example the Dollar has a 54% weight in the Mexican Peso NEER according to the BIS).

However, that is not always the case. For example, the Dollar accounts for only 13% of the Thai Baht NEER. So while the Thai Baht may be seasonally weak versus the Dollar in a particular month, the Thai Baht NEER may be seasonally strong that month if its other, more heavily-weighted constituent currencies, including the Chinese Renminbi and Japanese Yen, are even weaker versus the Dollar that month. Put differently if the Thai Baht tends to appreciate versus the Dollar in a particular month this may tell us more about the Dollar’s seasonality than the Thai Baht’s.

Ultimately the NEER provides a more accurate overall picture (than the exchange rate versus the Dollar or other individual currencies) of a country’s currency and trade competitiveness and of the risks of imported inflation/deflation. Therefore, central banks particularly in open economies (including the People’s Bank of China) arguably attach greater importance to the NEER when setting monetary policy (see PBoC likely to keep Renminbi on tight leash, 22nd October 2020, and Far more to Renminbi than USD/CNY cross, 8th December 2020). The Monetary Authority of Singapore (MAS) explicitly targets the Singapore Dollar NEER, keeping it in an undisclosed band.

A number of factors can drive monthly currency seasonality, including (sometimes well known) underlying seasonal patterns in a country’s balance of payments (e.g. exports/imports, tourism receipts, worker remittances and capital flows) which themselves could be due to the timing of public holidays and the depth of market liquidity. However, even well established seasonal patterns of monthly outperformance or underperformance can break down due to significant changes in government or central bank policies, electoral cycles, major domestic events (e.g. Brexit related development for Sterling, terrorist attacks), hard to predict “acts of god” (e.g. earthquakes) and of course once-in-a-lifetime “black swan” events such as the Covid-19 pandemic which has swept our planet since January.

 

Monthly currency seasonality: Down and out? 1

 

Covid-19 and global risk sentiment wrecked historical currency seasonality in 2020

Unsurprisingly, typical monthly seasonal currency patterns broke down in 2020, with a few exceptions. In particular, when global risk aversion peaked in March only a handful of currencies moved broadly in line with historical patterns according to our calculations (see Figures 1& 2).

  • The Philippines Peso NEER appreciated about 0.5% mom in March 2020, broadly in line with its March 2010-2019 average of +0.1% mom (a gap of less than 0.4 percentage points – see Figure 1).
  • The Taiwan Dollar NEER, which in the past decade has exhibited very modest monthly seasonality, appreciated 0.8% in March 2020, broadly in line with its 2010-2019 average of +0.3% mom (a gap of only 0.45pp).
  • The Hungarian Forint NEER depreciated 0.8% in March 2020, broadly in line with its March 2010-2019 average of -0.3% mom (a gap of only 0.5pp).
  • The Romanian Leu NEER appreciated 0.65% in March 2020 versus an average appreciation of 0.05% mom in 2010-2019 (a gap of 0.6pp).

 

Monthly currency seasonality: Down and out? 2

 

The Korean Won NEER depreciated less than 0.6% in March 2020 versus an average appreciation of less than 0.3% mom in 2010-2019 (a gap of about 0.8pp). Similarly the Swedish Krona NEER depreciated less than 0.7% in March 2020 versus an average appreciation of about 0.2% mom in 2010-2019 (a gap of about 0.9pp).

Other major currencies either:

  • Appreciated when in the past they had on average depreciated or been unchanged (CHF, CNY, DKK, EUR, JPY and USD). At the extremes, the Dollar, Euro and Japanese Yen NEERs appreciated respectively 5%%, 4.7% and 3.7% whereas in the past they had on average depreciated 0.1%, 0.4% and 0.5% respectively.
  • Depreciated when in the past they had on average appreciated or been unchanged (AUD, CAD, COP, CZK, IDR, INR, MXN, MYR, PLN, RUB, SGD, THB and ZAR). At the extremes, the Mexican Peso and Russian Rouble NEERs depreciated respectively 14.6% and 13.5% in March 2020 whereas in the past previous ten years they had on average appreciated 1.5% and 2.3% respectively.
  • Depreciated substantially more than they had done on average in the past decade (BRL, CLP, GBP, NOK, NZD and TRY). At the extremes, the Brazilian Real and Norwegian Krone depreciated respectively 9.2% and 8.9% whereas in the past previous decade they had on average depreciated 0.6% and 0.1% respectively.

 

Similarly, when global risk appetite peaked in June 2020 the vast majority of currencies deviated materially from their historical seasonal patterns (see Figures 1 & 3). Broadly speaking currencies which had appreciated rapidly in March, including the US Dollar, Japanese Yen and to a lesser extent the Chinese Renminbi, depreciated materially in June. The main exception was the Euro NEER which appreciated a further 0.8% (broadly in line with its historical average gain of +0.15% mom).

Conversely, many high-yielding Emerging Market (EM) currencies (BRL, IDR, MXN, RUB and ZAR) and more risk-sensitive developed market currencies (AUD, NOK) which had depreciated sharply in March appreciated materially in June, against seasonal historical patterns for the month of June (depreciation). A number of currencies (CZK and NZD) also appreciated far more in June than they had on average done in the past decade.

A number of currencies (DKK, EUR, GBP, INR, PHP, RON, SGD, TRY and TWD) admittedly moved broadly in line with historical patterns in June 2020 but the overall picture was one of currencies largely ignoring history and beating to the tune of resurgent global risk sentiment.

 

Monthly currency seasonality: Down and out? 3

 

Domestic factors re-established seasonal currency patterns in Q3, far less so in Q4

Historical seasonal patterns somewhat re-established themselves in Q3 (notably August). Figure 1 shows that for these three months there were fewer green and red cells and far more yellow cells. We partly attribute this to the Covid-19 pandemic having somewhat subsided over the summer, governments (partially or fully) unwinding national lockdowns and market, economic and political factors at a domestic level having once again played more of a role in driving asset prices, including currencies.

However, the resurgence in Covid-19 cases and deaths since then along with governments’ re-imposition of tighter social distancing measures and in many cases national lockdowns once again contributed to a breakdown in historical seasonal patterns, albeit not as marked as in March or June. In December 2020, only about 13 out of 31 major NEERs (CHF, DKK, GBP, IDR, JPY, KRW, MYR, NZD, RON, SEK, SGD, TWD and THB) moved broadly in line with their historical seasonal patterns (see Figures 1 & 4).

 

Monthly currency seasonality: Down and out? 4

 

Low-yielding emerging Asian currencies have stuck closest to historical seasonal patterns

As noted in Nominal Effective Exchange Rates: Monthly seasonal patterns (10th January 2019), emerging Asian currencies, particularly the low-yielding Chinese Renminbi, Korean Won, Philippines Peso, Singapore Dollar, Taiwan Dollar and Thai Baht, have historically exhibited only limited seasonality compared to most other EM currencies (see Figure 5). We attribute this in large part to Asian central banks’ ability and willingness to actively manage their currencies (including via FX intervention) in a bid to limit volatility and cap directionality.

Even throughout 2020 these currencies on the whole deviated far less from their monthly historical patterns than most other major EM and developed market currencies, which again we ascribe to the more interventionist stance of their central banks. This has been particularly true for the:

  • Korea Won, with the notable exception of October (see Figure 6);
  • Taiwan Dollar, as already noted above (see Figure 7); and
  • Singapore Dollar, unsurprisingly perhaps given that the MAS formally manages its currency NEER in a band (see Figure 8). The notable exception was the month of February with Singapore Dollar weakness due in part to the government imposing tough social distancing measures.

 

The Romanian Leu is the only other major NEER which recorded modest monthly changes in 2020 and which consistently traded broadly in line with historical seasonality (see Figure 9).

 

Monthly currency seasonality: Down and out? 5

 

Monthly currency seasonality: Down and out? 6

 

Monthly currency seasonality: Down and out? 7

 

Monthly currency seasonality: Down and out? 8

 

Monthly currency seasonality: Down and out? 9

 

January 2021 and beyond – Will seasonal patterns make a comeback?

About 15 NEERs traded broadly in line with historical patterns in January 2020 according to our estimates (see Figures 1 & 10). However, the Covid-19 pandemic was still in its infancy (the World Health Organisation only described the outbreak as a pandemic on 11th March 2020) and its impact on global economies and financial markets was arguably still modest. Our core scenario for January 2021 is that many currencies may continue to deviate materially from their 2010-2019 seasonal patterns.

 

Monthly currency seasonality: Down and out? 10

 

According to data for 2010-2019, the NEERs to record the strongest average rates of appreciation in the month of January were the:

  • Colombian Peso (+1.3% mom) which was equivalent to almost two standard deviations (i.e. the Peso’s appreciation was statistically significant) – see Figure 11. The Peso NEER appreciated in eight out ten months between 2010 and 2019 and its pace of appreciation January 2020 (+1.1%) was in line with history (see Figure 10).
  • Brazilian Real (+1.2% mom) or about 1.7 standard deviations. However, the Real’s performance has been slightly more erratic. It appreciated in only seven out ten months between 2010 and 2019 and depreciated about 1.3% in January 2020.

 

The NEERs to record the strongest average rates of depreciation in the month of January were the:

  • Turkish Lira (-1.3% mom) but this pace of depreciation was equivalent to less than one standard deviations.
  • Russian Rouble (-1.1% mom) but again this was equivalent to less than one standard deviations.

 

Monthly currency seasonality: Down and out? 11

 

  • The Swedish Krona NEER on average appreciated about 0.8% in the month of January in 2010-2019, or about 1.7 standard deviations. However it depreciated about 0.9% in January 2020.
  • The Indian Rupee NEER on average appreciated about 0.4% in the month of January in 2010-2019, or about one standard deviation. However it depreciated about 0.6% in January 2020.
  • The Euro NEER on average depreciated about 0.4% in the month of January in 2010-2019, or just over one standard deviation, and it also depreciated about 0.4% in January 2020.
  • The Philippines Peso NEER on average depreciated about 0.34% in the month of January in 2010-2019, or just over one standard deviations, and it depreciated about 0.6% in January 2020.

 

The Swiss Franc NEER on average appreciated about 0.8% in the month of January in 2010-2019, or about one standard deviation, and appreciated about 1.2% in January 2020 when as noted above the Covid-19 pandemic was in its infancy.

However, this historical outperformance is mainly due to the Swiss Franc NEER’s 8.2% appreciation in January 2015 (when the Swiss National Bank caved into strong appreciation pressures). Moreover, whilst alpine resorts in Switzerland are currently open to the general public (unlike in the rest of Europe) most foreigners face material travel restrictions as well as a domestic quarantine which will severely limit the number of tourists who are able and willing to holiday in Switzerland in coming weeks and months. As a result the Swiss Franc may appreciate far less in January 2021 than it has typically done in the month of January or even depreciate, in our view.

Finally we note that the Swiss Franc NEER is currently trading broadly in the middle of a narrow 1.6% range in place since 10th June according to our calculations (see Figure 12). The odds of the Swiss Franc appreciating or depreciating in January 2021 are pretty evenly balanced, in our view, and likely to be conditioned by global risk sentiment.

 

Monthly currency seasonality: Down and out? 12

 

Monthly currency seasonality: Down and out? 13