Correlation Analysis to Demonstrate Effectiveness of Quantitative Easing by Bank of England Upon the Exchange Value of Pound Sterling

Home»Correlation Analysis to Demonstrate Effectiveness of Quantitative Easing by Bank of England Upon the Exchange Value of Pound Sterling

1.      Quantitative Easing – What and Why

Quantitative Easing (QE) is a strategy adopted by central banks to achieve macro-economic objectives of inflation, growth and employment by tweaking money supply, i.e. injecting into or withdrawing money from the economy, bypassing the banking system. QE is resorted to after conventional methods such as interest and reserve rate are exhausted. Under QE, central banks purchase assets such as government debt “gilts” and high quality private debt, in exchange for cash for circulation in the system, resulting in the cascaded / multiplier effect of stimulating investment and spending, thereby generating growth and employment. QE also tend to influence currency exchange rates (Bank of England (1), 2012).  

Asset purchase actions are expected to increase their prices and lower yield, resulting in lower yield expectations in the market, reducing borrowing costs for all and stimulating overall demand for all asset classes as well as consumption spending  (Bank of England (2)).

One view of QE is central banks expanding their balance sheet through asset purchases, financed by central bank money without printing of currency (Bernanke and Reinhart, 2004).

The distinction between QE and printing money needs mention. While QE is a funds deployment tactic of central banks, printing currency implies creation of money to finance government deficits or retire government debt (also known as monetizing the government debt). QE does not increase aggregate money supply, but only the structure of the money supply and is to be seen as temporary measure without any potential long term damage.

On the exchange rate front, enhancing money supply leads to disequilibrium in the demand supply equation tending to depreciate own currency against other currencies. This could be interpreted as a conscious move to benefit exporters from the country enhancing their competitiveness and debtors whose real value of repayment depreciates; and potentially harming creditors holding the currency. Fall in currency valuation is detrimental to importers by escalating their cost in home currency terms.

 2.      Analysis Done

This paper analyzes the relation between QE and Exchange rates between the sterling pound and the US Dollar over the period 12th March 2009 to 29th September 2011 (134 weeks). The QE activity, during the period is characterized by two distinct contiguous periods of easing / expansion (42 weeks) and inaction / contraction (84 weeks).  Eight weeks experienced a mix of expansion and contraction.

Following analysis was done

Correlation for the entire 134 weeks and two sub-periods of expansion and contraction, under two scenarios: (a) concurrent variation in exchange rates and (b) variation in exchange rate with a one week lag to the QE activity.

Correlation using data on cumulative QE and incremental QE under the above six scenarios

Regression analysis of the QE against week as independent variable

 The above approach was adopted to explore existence of varying behavior patterns during expansion and contracting phases (http://www.bankofengland.co.uk). This approach helped validate the expectation theory on market behavior (The United Kingdom’s quantitative easing policy: design, operation and impact – (By Michael Joyce, Matthew Tong and Robert Woods of the Bank’s Macro Financial Analysis Division in Quarterly Bulletin 2011 Q3, Research and analysis The United Kingdom’s quantitative easing policy).

Correlation analysis structure is depicted as below

1.      Data Source

 

  1. Quantitative easing for the period March 2009 to September 2011 - http://www.theguardian.com/news/datablog/2011/oct/04/quantitative-easing-bank-england# zoomed-picture

 

  1. for exchange rate – http://www.bankofengland.co.uk/boeapps/iadb/fromshowcolumns.asp?filter=n&travel=nixscx&shadowpage=1&searchtext=xudluss%2cxudlers%2cxudlsfs%2cxudlbk85%2cxudlads%2clpmvqjw%2cxudlbk95%2cxudlsks%2cxudlcds%2cxudlsgs&searchexclude=&searchtextfields=tc&thes=&searchtype=&cats=&actualresnumperpage=&totalnumresults=10&xnotes=y&c=c8p&fnotes2=y&xnotes2=y&showdata.x=51&showdata.y=7

 

2.      Data Analysis and findings

 

  • Concurrent data on QE and exchange rates

The correlation between QE on an incremental basis and exchange rate data for the whole period (134 weeks) comprising expansion, quiet and contraction doesn’t indicate any credible relation (0.054). The same exchange rate data against cumulative holding (comprising expansion and contraction) shows a low level of +ve correlation (0.239). A positive correlation is to be expected as  higher the asset holding, more cash is released into the system and a higher value of exchange rate indicates a depreciating currency (more of USD for same £). Thus the findings corroborate the theory that more of domestic currency supply into the system tends to devalue it against a foreign currency.

Findings from the same analysis done for the contiguous monetary expansion period of 42 weeks indicate a strikingly high positive correlation of 0.819 between QE and exchange rate. One has to infer that a contiguous period of consistent expansionary policy sets clear expectation in the market, leading to observed high correlation between monetary expansion and a weakening currency. A similar reverse phenomenon is visible for the contiguous 84 weeks period of contraction at -0.710, though at a marginally lower level. When we examine the correlation of exchange rates against incremental values of QE, the correlations are of lower value (-0.331 and +0.2) and also surprisingly unexpected sign, both on purchase (expansionary) period as well as the sale (contraction) period (Table1). Since the analysis is on data for the same day, it is to be inferred that (a) cumulative values are more influential in shaping market behavior, implying expectation setting, memory and implied learning effect on policy direction over market players.

Table 1 - Correlation Analysis  Between QE Action and Exchange Rate on Same Day

No. of observations

134

134

42

84

42

84

 

Variables

Incremental asset change & exchange rate

Cumulative asset holding & exchange rate

Cumulative asset purchased & exchange rate

Cumulative asset sold & exchange rate

Incremental asset purchased & exchange rate

incremental asset sold & exchange rate

 

 

             

 

 

Correlation 

co-efficient

0.054

0.239

    0.819

-0.710

-0.331

0.200

 

                                                       

 Exchange rate data lagging QE by one week

In this scenario there are 133 (one less than in the concurrent analysis), 42 and 84 data pairs under the whole, expansion and contraction period respectively. The incremental and the cumulative data for the whole period (133 data pairs) indicate low positive correlations (0.112 and 0.162), while the expansionary and contraction period under cumulative scenario demonstrate distinct correlation in the expected direction (0.786 and -0.723), that is comparable to the concurrent data scenario (above). This implies that a lag of one week is too short a period for the market (indifferent) to take a different view on central bank’s direction / expectation on policy. Examination of results based on the incremental assets purchased / sold has thrown up values comparable to the earlier scenario (-0.325 and +0.199), both in value and direction. Incremental asset transactions may be seen as sending confusing signals, and what matters is the demonstrated consistency in policy signals through visible actions (Table 2). 

 

 

Table 2 - Correlation  Coefficient With One Week Lag for Exchange Rate

 

o. of observations

133

133

42

84

42

84

 

Variables

Incremental asset change & exchange rate

Cumulative asset holding & exchange rate

Cumulative asset purchased & exchange rate

Cumulative asset sold & exchange rate

Incremental asset purchased & exchange rate

Incremental asset sold & exchange rate

             

Correlation  coefficient

0.112

0.162

0.786

-0.723

-0.325

0.199

 

Regression equations on cumulative asset holdings against weeks under three scenarios: whole (R² = -0.51), expansionary (R² = +0.967) and contraction (R² = +0.908) periods demonstrate a definite direction only for the expansion and contraction periods treated independently, for which periods the correlations with exchange rates are also decisive, as seen in both scenarios above (Exhibits A,B and C). The regression on incremental assets show a poor trend as is to be expected since these are independent transient decisions over time (Exhibits D, E and F).

Exhibit A – Cumulative holding of assets for 134 weeks

Exhibit B – Cumulative Asset Purchases for Expansion Period of 42 Weeks

Exhibit C- Cumulative Asset Sold for Weeks 50 To 133

Exhibit D – Incremental Purchase of Assets for Whole Period of 134 Weeks

Exhibit E –Incremental Asset Sales for Contraction Period Weeks 50 to 133

(d) Regression Analysis of Exchange Rates and Expansionary / Contraction actions

Six linear regressions have been run with Exchange rate as dependent variabe and asset holding cumulative, incremental purchase and sales for both current period exchange rate as well as one week lagged period. Barring the cumulative asset holding during expansionary and contract period, all other correlations are very low (low R² values), indicating  poor fit (Exhibits G,H,I J,K,L).

Exhibit G – Exchange Rate Current Period – Cumulative Asset Holding 134 Weeks

Exhibit H – Exchange Rate Curent Period – Cumulative Asset Holding Expansionary 42 Weeks

Exhibit I – Exchange Rate Current Period – Cumulative Asset Holding Contraction – 84 Weeks

Posted by: Kooveli Madom. in Finance | Date: 13/02/2016

Share this article

Other articles

Back to article listing

Find a Great Teacher

Tell us your learning needs in detail and get immediate response from qualified tutors

Terms & conditions agreed
Ask a Question
Top