The (Divisia) money trail – a very bullish UK story

Recently, the data for the UK economy has been very strong, and it is very clear that the UK economy is in recovery. So what is the reason? Well, you guessed it – monetary policy.

I think it is fairly easy to understand this recovery if we follow the money trail. It is a story about how UK households are reducing precautionary cash holdings (in long-term time deposits) because they no longer fear a deflationary scenario for the British economy and, that is due to the shift in UK monetary policy that basically started with the Bank of England’s second round of quantitative easing being initiated in October 2011.

The graphs below I think tells most of the story.

Lets start out with a series for growth of the Divisia Money Supply in the UK.

Divisia Money UK

Take a look at the pick-up in Divisia Money growth from around October 2011 and all through 2012 and 2013.

Historically, UK Divisia Money has been a quite strong leading indicator for UK nominal GDP growth so the sharp pick-up in Divisia Money growth is an indication of a future pick-up in NGDP growth. In fact recently, actual NGDP growth has picked up substantially, and other indicators show that the pick-up is continuing.

If you don’t believe me on the correlation between UK Divisia Money growth and NGDP growth, then take a look at this very informative blog post by Duncan Brown, who has done the econometrics to demonstrate the correlation between Divisia (and Broad) Money and NGDP growth in the UK.

Shifting money

So what caused Divisia Money growth to pick-up like this? Well, as I indicated, above the pick-up has coincided with a major movement of money in the UK economy – from less liquid time deposits to more liquid readably available short-term deposits. The graph below shows this.

Deposits UK

So here is the story as I see it.

In October 2011 (A:QE in the chart), the Bank of England restarts its quantitative easing program in response the escalating euro crisis. The BoE then steps up quantitative easing in both February 2012 (B: QE) and in July 2012 (C: QE). This I believe had two impacts.

First of all, it reduced deflationary fears in the UK economy, and as a result households moved to reduced their precautionary holdings of cash in higher-yielding time deposits. This is the drop in time deposits we are starting to see in the Autumn of 2011.

Second, there is a hot potato effect. As the Bank of England is buying assets, banks and financial institutions’ holdings of cash increase. As liquidity is now readily available to these institutions, they no longer to the same extent as earlier need to get liquidity from the household sector, and therefore they become less willing to accept time deposits than before.

Furthermore, it should be noted that in December 2012, the ECB started its so-called Long-Term Refinancing Operation (LTRO), which also made euro liquidity available to UK financial institutions. This further dramatically helped the liquidity situation for UK financial institutions.

Hence, we are seeing both a push and pull effect on the households’ time deposits. The net result has been a marked drop in time deposits and a similar increase in instant access deposits. I believe it has been equally important that there has been a marked shift in expectations about UK monetary policy with the appointment of Mark Carney in December 2012 (D: Carney).

Mark Carney’s hints – also in December 2012 – that he could favour NGDP targeting also helped send the signal that more monetary easing would be forthcoming if needed, as did the introduction of more clear forward guidance in August 2013 (E: ‘Carney Rule’). In addition to that, the general global easing of monetary conditions on the back of the Federal Reserve’s introduction of the Evans rule in September 2012 and the Bank of Japan’s aggressive measures to hit it new 2% inflation undoubtedly have also helped ease financial conditions in Britain.

Hence, I believe the shift in UK (and global) monetary policy that started in the Autumn of 2011 is the main reason for the shift in the UK households’ behaviour over the past two years.

Monetary policy is highly potent

But you might of course say – isn’t it just money being shifted around? How is that impacting the economy? Well, here the Divisia Money concept helps us. Divisia money uses a form of aggregation of money supply components that takes this into account and weights the components of money according to their usefulness in transactions.

Hence, as short-term deposits are more liquid and hence readably available for transactions (consumption or investments) than  time deposits a shift in cash holdings from time deposits to short-term deposits will cause an increase in the Divisia Money supply. This is exactly what we have seen in the UK over the past two years.

And since as we know that UK Divisia Money growth leads UK NGDP growth, there is good reason to expect this to continue to feed through to higher NGDP growth and higher economic activity in Britain.

Concluding, it seems rather clear that the quantitative easing implemented in 2011-12 in the UK and the change in forward guidance overall has not only increased UK money base growth, but also the much broader measures of money supply growth such as Divisia Money. This demonstrates that monetary policy is highly potent and also that expectations of future monetary policy, which helped caused this basic portfolio readjustment process, works quite well.

“Monetary” analysis based on looking at interest rates would never had uncovered this. However, a traditional monetarist analysis of money and the monetary transmission mechanism, combined with Market Monetarist insights about the importance of expectations, can fully explain why we are now seeing a fairly sharp pick-up in UK growth. Now we just need policy makers to understand this.



I think some acknowledgements are in place here as this blog post has been inspired by the work of a number of other monetarist and monetarists oriented economists and commentators. First of all Britmouse needs thanking for pointing me to the excellent work on the “raid” on UK households’ saving by Sky TV’s economics editor Ed Conway, who himself was inspired by Henderson Economics’ chief economist Simon Ward, who has done excellent work on the dishoarding of money in the UK. My friend professor Anthony Evans also helped altert me to what is going on in UK Divisia Money growth. Anthony himself publishes a similar data series called MA.

Second of course, a thanks to Duncan Brown for his great econometric work on the causality of Divisia Money and NGDP growth in the UK.

And finally, thanks to the godfather of Divisia Money Bill Barnett who nearly single-handledly has pushed the agenda for Divisia Money as an alternative to simple-sum monetary aggregates for decades. In recent years, he has been helped by Josh Hendrickson and Mike Belongia who has done very interesting empirical work on Divisia Money.

For a very recent blog post on Divisia Money, see this excellent piece by JP Koning.

And while you are at it, you might as well buy Bill Barnett’s excellent book “Getting It Wrong” about “how faulty monetary statistics undermine the Fed, the financial system and the economy”.


Leave a comment


  1. Great post Lars! Ive been very enthusiastic regarding DIVISIA measures since reading Ireland/Belongia paper on money/interest relation to real variables (recommended by D.Beckworth in one post). This looks like a good story for the UK! There are no DIVISIA indexes for Euro area yet?

  2. marksadowski

     /  December 9, 2013

    The Duncan Brown results are interesting. However, I caution him that Granger causality tests should always be done on level data otherwise the results will be inconsistent. This is discussed in detail by James Hamilton in “Time Series” on pages 651-3. In particular, all of my Granger causality tests are done in levels using a VAR technique developed by Toda and Yamamato that addresses the issue of nonstationarity by adding additional lagged terms as exogenous variables.

    I’ve conducted Granger causality tests on the US and found bidirectional Granger causality between the monetary base, M1, M2, MZM and NGDP on quarterly data since 1959 all at the 1% significance level.

    A couple of weeks ago I expanded my Granger causality tests to more thoroughly cover five zero lower bound episodes when the monetary base was significantly expanded (i.e. when there was QE): 1) US from May 1933 to February 1937, 2) US from July 1937 through November 1941, 3) Japan from March 2001 through May 2006 4) the US from December 2008 through September 2012 and 5) the UK from April 2009 through September 2013. Note that since Japan only started QE in earnest in April 2013 I have not included this episode yet because the period is too brief. The time periods were selected based on the level of interest rates. The US periods in the 1930s are when the 3-month T-Bill yield was 0.29% or less. The Japanese period was selected based on when the call rate was 0.1% or less. The most recent US episode is when the federal funds rate dropped below 0.25%. The UK period was selected based on when the bank rate was 0.5% or less.

    All of the results were interesting, particularly the US results, but I’ll only report my UK results here as they are most relevant.

    The UK’s measure of broad money is denoted M4. In addition the BOE also estimates M1, M2 and M3 measures according to Euro Area standards. The lending counterpart of M4 is denoted M4Lx. In addition I ran tests on Retail M4 (retail deposits only), M4ex (without other financial corporations) and the lending counterpart of M4ex which is M4xex.

    I find Retail M4 Granger causes M4lx at the 10% significance level, M1 Granger causes M4Lx at the 10% significance level, M4 Granger causes M4Lxex at the 10% significance level and M4Lxex Granger causes M4ex at the 5% significance level. The first three results run counter to Accomodative Endogeneity.

    The monetary base Granger causes 5-year inflation expectations as measured by inflation-linked gilts at the 1% significance level. The monetary base Granger causes the real effective exchange rate of the pound sterling and the industrial production index at the 5% significance level. The monetary base Granger causes the deposit component of the Euro Area measures of broad money at the 10% significance level. In addition the deposit component of M4 Granger causes the monetary base at the 10% significance level.

    I evidently need to look at Divisia measures next.

    • marksadowski

       /  December 9, 2013

      “4) the US from December 2008 through September 2012”

      should read

      “4) the US from December 2008 through September 2013”

    • Thanks Mark…that is extremely helpful.

      I need to convince you to write a guest post on your results!

    • Mark: the VARs were done in levels form, of the log of the money/credit aggregate along with the log of nominal GDP (all the data go from Q1 1977 to Q2 2013).

      As you’re a more discerning customer, I had a go at a T-Y procedure and get fairly similar results. The money > spending causation remains 1% significant for both of the Divisia indices, while M4 households and corporates and both M4L indices all now show causation running spending > money. So Divisia holds up pretty well, regardless. In terms of accommodative endogeneity, Divisia Granger causes M4L (and not vice versa) at the 1% significance level, but most other measures are (as you would expect) much messier.

      Lars: great post, especially the chart with the timings of policy moves, and thank you for noticing my own post too.

  3. Re the decline in time deposits, it is worth noting that the rate paid on such deposits fell significantly over roughly the same period as the decline in holdings shown in your graph.

  4. Yes, the ECB has a Divisia monetary aggregates database. I set it up for them. You can find my working paper in the ECB working paper series. For that purpose, I had to develop the extension to multilateral monetary aggregation, so they could first aggregate within countries and then over countries. The working paper subsequently was published in shorter form by the J. of Econometrics:

    “Multilateral Aggregation-Theoretic Monetary Aggregation over Heterogeneous Countries,” Journal of Econometrics, vol 136, no 2, February 2007, pp 457-482.

    The ECB makes that data available to its Governing Council, when it meets. I don’t know why they do not make the data public.

    The “Godfather” 🙂

    • Thanks for the input “Godfather”;-)

      It says a lot about the ECB that the bank continue to not make this data available to the public. The ECB do not want public debate about the quality of policy making in the euro zone. That is too bad…

      • William A. Barnett

         /  December 9, 2013

        Sadly the ECB is not the only central bank that does that. For example, the Bank of Japan was one of the first central banks to adopt Divisia monetary aggregates, shortly after I spoke at a conference in Tokyo long ago. I have never seen that data. The only reason I know about it is because they hired Allan Meltzer as a consultant to help them with it.

  5. oldcobbler

     /  December 10, 2013

    “UK households are reducing precautionary cash holdings (in long-term time deposits) because they no longer fear a deflationary scenario for the British economy”.

    If by “deflationary scenario” you mean a falling price level, did UK households ever fear a deflationary scenario for the British economy ?

    Page 35 of the latest Bank of England inflation report ( ttp:// in fact suggests households’ inflation expectations have fallen between 2011 and now.

  6. RaviVarghese

     /  December 10, 2013

    I’ve only encountered Divisia in passing but this has inspired me to read Barnett’s book (along with a guest appearance from The Godfather himself!). However, I think there’s something a bit counterintuitive here. I generally think of money demand as increasing when people shift from less liquid assets (stocks, bonds) to more liquid (deposits). Here we see the opposite – people going from less liquid to more liquid but the demand for money DECREASING. Perhaps this is just a flaw in my prior framework, or an advantage of the Divisia framework that it shows more clearly changes in precautionary and transaction demand.

  7. Very interesting post Lars!

    I do have exactly the same problem as Ravi above though, that if we want to explain 2008 as a money demand shock this data does *not* appear to be supportive, because households were dumping liquidity in 2008/9. Particularly this looks like the opposite of what happened in the US…

    A lot to think about here.

  8. Sentient Being

     /  December 11, 2013

    An anecdote from our household. We were not convinced that the financial system was going to survive during 2009. We feared the introduction of depositor confiscation (“bail-ins”, or “Cyprussing” as I now call it). We feared that one day we might be told that our money in the banking system was still there, and still ours, but we wouldn’t be allowed to withdraw it. We tried withdrawing cash but met with stiff resistance and some very hostile questioning from the banks and building societies. It was quite remarkable how the attempted withdrawal in cash of amounts such as £5,000 during 2009 was met with reticence, evasion, and delay, This made us even more fearful. Our reaction was to cease paying cash into the bank from our cash-generative business. We used funds from maturing term deposits to cover cheques to buy the cash from our business, obviously accounting for these transactions properly. We accumulated £30,000 in cash from this, all in small denomination banknotes which we stored in a safe place. During 2011 and 2012 we decided to invest this cash but because of anti-money laundering rules we had to move it back into the banking system first. I wonder whether others were behaving similarly during this period?
    Our current behaviour is a determined effort to get rid of all our cash and all deposits in the financial system. We have a line of credit in the form of a partly used offset mortgage and we can draw on the balance at any time so this gives us a “buffer” if we get our sums wrong and take a bit too much out at any time. We have nearly completed the closure of all our savings accounts (except for an allegedly inflation-proof NS&I bond). The money is now invested in a spread of assets which we think will survive a period of rapid reduction in the purchasing power of paper money, and a temporary seizure of the UK financial system. Our current tax position means that we are not motivated to chase yield like so many and therefore our primary objective is capital protection. We have been very influenced by Mark Carney’s track record in Canada and we did quite a bit of research into what he did there and what the results were. Of great interest to us is our observation that all the people we know, without any exceptions whatsoever, are either wealthy (and nearly all of those are of similar mature age to ourselves) or living payday to payday with negligible net liquidity. We don’t know any households who we could call examples of the traditional “middle class” with a regular surplus of income who are willing and able to save money in the traditional way over and above their pension contributions. Not one.
    As I said, just an anecdote from one household.

  9. William A. Barnett

     /  December 11, 2013

    Ravi: I am not sure that I understand your hypothesis, and I am not an expert on the British economy. But, yes, a transfer of funds from less liquid assets to more liquid assets would increase the monetary service flow and thereby the broad Divisia monetary aggregate — if all else is constant, including all interest rates, prices, and total portfolio expenditure. Of course one would then be left with the problem of explaining the transfer, when all relevant explanatory variables had not changed. For example, if the substitution occurred as a result of a Hicksian compensated change in relative interest rates, then monetary services and thereby the Divisia monetary aggregate would be constant. In short, your question is more complicated than it might have appeared to be.

    Most misunderstandings about the Divisia monetary aggregates result from falling into the trap of one of the well-known paradoxes in microeconomic theory, such as the famous “diamonds versus water paradox.” I hope that your question will motivate you to learn the relevant theory, so you will be able to answer such questions on your own.

    The best place to look is Appendix E (“Understanding Divisia Aggregation”) in my book, “Getting It Wrong.” But if you would prefer to start with some lighter reading before taking that plunge, I recommend my chapter in the forthcoming book about Milton Friedman. The working paper version is online at:

    Click to access Friedman%20and%20Divisia%20Monetary%20Measures.pdf

    If you get really good at it, I might appoint you my Consigliere, 🙂

    The Godfather

  1. Is this a booming economy, or a credit-driven bubble? – Telegraph Blogs
  2. The massively negative euro zone ‘Divisia Money Gap’ | The Market Monetarist

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: