Libor manipulated? Yep. What else?
Posted on 24 August 2012 by
The following article first appeared in last week's edition of Financial Adviser:
When the Treasury select committee quizzed Sir Mervyn King and Lord (Adair) Turner recently about the Libor rate fixing scandal, both had their excuses ready in response to one of the obvious questions - why did they take no action when told by the New York Fed that some UK banks were falsifying Libor submissions? “Not my responsibility, Guv,” they chimed in unison.
Both claimed it was the responsibility of the British Bankers’ Association, as Libor was its baby, although in addition Sir Mervyn flagged up that the FSA was the banking regulator - not the Bank. Passing the buck to the BBA might have been justified if banks were self-regulating. Silly me but for all these years I have been labouring under what must be a misapprehension that the previous government gave the FSA statutory responsibility to regulate banks.
In fact, the Market Conduct Sourcebook in the FSA Handbook says that regulated markets are obliged to ensure “fair and orderly trading, governed by transparent, non-discretionary rules.” Presumably Lord Turner would acknowledge that part of a regulator’s job is to ensure its rules are not broken.
Perhaps he would explain how trading in the hundreds of trillions of pounds of Libor-based contracts can be “fair” when Libor rates are being deliberately falsified. Likewise, how the “non-discretionary rules” requirement was adhered to when Libor rate submitters made up rates as they went along, rather than submitting their view of the actual rate as required by BBA rules.
Clearly the primary responsibility for the Libor manipulations must be with the banks who submitted false rates. But how extraordinary that when the Bank of England is warned by what is probably the most powerful regulator in the world that some UK banks are making fraudulent Libor submissions, and passes that information to the FSA, it is completely ignored by the latter and apparently not followed up by the former.
The only implication I can draw from this is that either the FSA, and to a lesser extent the Bank of England, condoned the fraud and were thus complicit in it, or they were grossly incompetent.
The manipulation of Libor begs the question about what other figures are being manipulated, not necessarily fraudulently. In fact, most regular monthly statistics on the housing and mortgage market are manipulated, or to use the technical term, seasonally adjusted, with the Bank of England being one of the biggest culprits.
Now of course seasonal factors do affect some statistics but so do many other things. For example mortgage availability obviously has a major impact on mortgage approvals and mortgage lending. This and many other factors, such as interest rates and the state of the economy, affect house prices, some much more than seasonal factors do.
So if one is going into the manipulation game why only manipulate house prices by the inexact art of seasonal adjustments? Why not, for example, adjust monthly house price statistics to reflect variations in interest rates? Pick a number that whoever is manipulating the statistics thinks is a “normal” interest rate. Then when rates are below that level reduce the actual monthly house price index before publishing it on the basis that the low interest rate has artificially inflated prices, and vice versa.
Why not also adjust for variations in all the other things that affect house prices, including inflation, and in the unlikely event of the economist who devised all the manipulations getting all the calculations correct house prices would remain unchanged forever.
You are probably thinking, what a stupid idea, and of course it is. But it is no more stupid than seasonal adjustments.
To make matters even worse many statistic compilers, with the Bank of England and the Land Registry being two of the least transparent, make such limited reference when releasing statistics to the fact that they have been manipulated by seasonally adjusting them that most reports of the figures completely ignore this fact and hence are misleading.
Because the Bank’s press releases do not include the genuine figures the only way to obtain them is to delve into its website. But although this has masses of information it sometimes does not include the genuine figures until some time after the manipulated ones have been released. However, even late publication is better than the service provided by the Land Registry, which cannot be bothered to publish the actual figures at all.
I first became dubious of seasonal adjustments in the 1970s, while working as a stockbroker in the gilt edge and fixed interest department. The monthly trade figures used to have a much bigger impact on the market than they do now and were announced, always only on a seasonally adjusted basis, by the government department then known as the Board of Trade at 3.30pm on a Friday. (In those days the gilt market closed at 3.30pm to allow the figures to be assimilated over the weekend before the market re-opened.)
One Friday after the figures had been announced, our chief economist called the Board of Trade at about 3.45pm, and said: “I’ve seen the seasonally adjusted trade figures but could you please tell me what the unadjusted figures were.” After a short pause the answer came back: “Oh, we haven’t worked them out yet.”
One of the few examples where statistics from different compilers are broadly designed to measure the same thing on the same timescales are the Nationwide and Halifax house price indices. In both cases the figures highlighted and commented on in the press releases are the manipulated, or seasonally adjusted, ones. However the actual figures are readily available, with Nationwide commendably including them in its monthly press releases.
Logic would suggest that even if the actual figures from these two sources diverged significantly, as often happens on a monthly basis, the monthly seasonal adjustments should nevertheless always be in the same direction and of a similar amount. However, often one lender’s seasonal adjustment increases the actual figure and the other’s reduces it. Even in the months when the seasonal adjustments are in the same direction the size of the adjustment often varies considerably.
One of the few seasonal factors where one can be fairly confident of an impact is the number of working days in a month. In June because of the double bank holiday, compared with a single day in May last year, plus the incidence of weekends, there were 19 working days in 2012 compared to 22 in 2011 – 13.6 per cent less. That obviously affects the volume of activity in things that happen primarily only on normal working days, such as processing mortgages, and hence was no doubt a major factor in the weak mortgage numbers reported for June by the Bank of England.
However, when reporting June’s mortgage lending figures the Bank said “the seasonal adjustment process has not automatically taken this effect into account,” (referring to the two extra bank holidays, not the extra weekend day). So the Bank ignores one of the few robust seasonal factors but manipulates figures using someone’s guesstimate of other seasonal factors.
The time is well overdue for all statistics to be reported using actual figures, not after being gerrymandered by a seasonal adjustment manipulation.
The blog postings on this site solely reflect the personal views of the authors and do not necessarily represent the views, positions, strategies or opinions of John Charcol. All comments are made in good faith, and neither Charcol Limited nor Ray Boulger will accept liability for them.