Sunday, November 28, 2010

Don Giovanni, Penrose and Equivalent Expected Cost

Turned on Radio 3 last night before our dinner party and caught the first part of the ENO Don Giovanni. As we suspected, Ian Patterson sings Don G wonderfully, and Jeremy Sams' very free translation contains many gems. But some of the other singers rather grated, and I don't like what I read about the staging at all. In any case I don't think there will be time to get along to it: but the next time Patterson sings I definitely want to be there.

We put on our amazing recording which has Schwartzkopf as Donna Elvira and Sutherland as Donna Anna: surely two of the greatest, if not the two greatest, sopranos of the late 20th C. I heard both of them, and indeed met Sutherland briefly after her last performance at Covent Garden. How many other recordings have both of them I wonder? This recording was made in 1959 when Schwartzkopf was 44 and Sutherland 33.

I've read Gurzadyan and Penrose and it is extremely interesting. I suspect it will be as controversial as Nowak, Tarnita and Wilson, since it calls into question a roughly similar piece of scientific orthodoxy. On the other had, the mathematics of Inflation is pretty well worked out, it is just that the physics is missing, whereas the maths of Extended Fitness is basically only a rough approximation, albeit one which has provided a useful experimental framework. The problem of course is that it is very difficult to be sure how likely these concentric circles would be:

a. To occur by chance.
b. To occur in some well-defined inflationary model.

If course if we had pure Guassian data we could work out (a) fairly simply by monte-carlo methods. But these cosmological maps have been subject to very heavy signal processing, which makes it much more difficult. And as for (b) I don't think people have investigated this question: furthermore there are a large number of inflationary models and plenty of parameters to tweak, even before we get into String/M theory. It should certainly set the agenda for some very interesting research.

I had a really promising idea provisionally dubbed the Principle of Equivalent Expected Cost yesterday. We'll see how it develops. And I just came across a lovely Alex cartoon which captures one of the points made in Ned's conference beautifully.

Friday, November 26, 2010

Science, Monty, Churchill and HYLAS

Lots happening mostly unbloggable.

Mon and Tues was at a conference and met a number of very interesting people, including John Beddington and Alice Huang. A couple of policy initiatives from this conference may become bloggable later.

Met Viscount Montgomery of El Alamein on Tues evening, and told him that his father had taken the salute at our march-past when I was at Prep School. Apparently his dad rather enjoyed doing that. I had also known slightly Carol Mather who had been my MP. Apparently, although it was true that Mather didn't know about Ultra, he did know that Monty was getting some very secret intelligence, but not how it was obtained.

Almost finished the Jenkins book on Churchill. During the end of his final premiership he was very focused on avoiding nuclear war. One of his most memorable late turns of phrase was "when the advance of destructive weapons enables everyone to kill everyone else, no one will want to kill anyone at all. At any rate, it seems pretty safe to say that a war which begins by both sides suffering what they dread most...is less likely to occur than one which dangles the lurid prizes of former ages before ambitious eyes" I wish I'd known this quote earlier, I'd have used it in Beale & Pollack. And indeed this one: "What ought we to do? Which way shall we turn to save our lives and the future of the world? It does not matter so much to old people; they are going soon anyway; but I find it poignant to look at youth in all its activity and ardour and, most of all, to watch little children playing their merry games, and wonder what would lie before them if God wearied of mankind"

Watched the launch of HYLAS I this evening - a really interesting initiative and the first public-private partnership of the ESA. Have also been reading the excellent McKinsey report From Austerity to Prosperity.

Tuesday, November 23, 2010

Columbia Conference, part 2


Richard Nelson then spoke about some ideas around evolutionary economics. And Phillipe Aghion gave a very exciting presentation suggesting, with a lot of convincing empirical data, that the benefits of counter-cyclical policy interventions in recessions were concentrated on those sectors which would otherwise spend on innovation but which are credit constrained. This has extremely important policy implications: it suggests that macroeconomic countercyclical interventions are very inefficient and that interventions should be focused on removing the liquidity constraints that innovation-intensive industries face.

Then Katarina Juselius used detailed analysis of the swings from PPP exchange rates in the DM/$ to demonstrate that the deviations were more an I(2) process than an I(1) which is a good argument against Rational Expectations Hypothesis and in favour of something like Imperfect Knowledge Economics, though I think evolutionary models could give similar results.

At lunch we had delightful speeches from Dale Mortensen and Chris Pissarides.
Gylfi Zoega then gave an intriguing paper, co-authored with Ned, about the relationship between Employment, Investment and Asset Prices, showing very convincing data that share prices and investment are good predictors of employment. By using 5-year data comparing share prices (normalised by labour productivity) in years 1-3 with employment in years 3-5 they are able to establish some causality, although as the discussant pointed out share prices are themselves estimates of future conditions so the underlying driver may well be confidence about future business prospects.

Roger Farmer followed with an argument for introducing “Belief Functions” to correct some of the clear defects of the neo-Keynsian REH view. John Taylor then discussed very convincingly the value of rules-based policy responses, suggesting inter alia that the departure from the Taylor Rule in 2002-7 was at least in part responsible for the asset price bubbles. He also demonstrated that the 2008 and 2009 stimulus packages had very little effect on personal consumption expenditures. Andy Haldane gave a superb paper on Curbing the Credit Cycle, noting that the Credit Cycle was not the same as the business cycle, that exaggerated credit cycles tended to lead to financial crises, and that macro-prudential tools would be highly desirable in helping to curb such cycles. And Michael Woodford argued for Principled Policy Making in an Uncertain World. Amar Bhide in his discussion was rather negative about the possibility of Macro-prudential regulation: I wonder whether he fully understands that it is like prose: even a decision not to regulate is a form of regulation, just a very bad one. I also worry about comparisons between US banking from 1950-1980 and now that doesn't seem to recognise that it is much easier regulating a system where the US is growing substantially and is a major creditor nation to one where it is very highly indebted.

Finally Gillian Tett gave a keynote speech at the dinner which was a masterly comparison of finance and her experiences as an anthropologist, drawing attention both to the dangerous siloisation in banks and indeed in society in general, and to the importance of social silences.

Altogether a fascinating and highly productive couple of days, and one in which the policy implications are already, to my certain knowledge, being drawn to the attention of relevant minsters.

Sunday, November 21, 2010

Columbia Conference - Part 1

When I had lunch with Ned last Friday he invited me to his amazing conference on Microfoundations of Modern Macroeconomics on the occasion of the 40th anniversary of the seminal conference in Pittsburgh that led to his Nobel Prize. I managed to clear the diary and flew in on Thurs evening, writing this on the plane back.

The initial presentations by Evans & Honkapohja, and Guesnerie talked about learning processes and market failures. They emphasise the most fundamental aspect in which Economics differs from the natural sciences which is the role of expectations. However the key theme of this conference is that the so-called “Rational Expectations Hypothesis” is fundamentally wrong and needs to be replaced, though it is not yet clear quite how to do this. Sheila Dow reflected on Keynes, noting that he began as a philosopher with a treatise on probability. Then Roman Frydman presented his ideas on Imperfect Knowledge Economics (IKE). It seems to me that their critique of REH is spot-on but that they haven’t yet developed a really convincing way forward.

Franklin Allen then discussed the role of central banks and property booms with a simple model in which speculators who could borrow cheaply caused bubbles. This was very interesting, though it raised the question of why banks are willing to lend to speculators at substantially below the risk-adjusted rate of return. Frydman and Goldberg then indicated how IKE might address this problem, with the introduction of an asymmetric utility function, so that actors care more about avoiding losses than making profits, and emphasising the important point that people’s expectations of future events vary: indeed it is this that drives markets in most cases. They use analysis of Bloomberg Market Wrap stories to demonstrate that the market is more driven by fundamentals than psychology: though the fact that journalists report that “the market was up because of good news from X” doesn’t necessarily prove that sentiment was not the reason, merely that it was not considered acceptable to say that sentiment was the reason.

Blake LeBaron then presented some interesting agent based simulation work in which fund managers with different learning approaches caused asset price behaviour similar to that observed. Not only do the managers update their diverse strategies, but money flows to the successful managers. This work is very promising, incorporating the elements of imperfect knowledge, diverse strategies and evolutionary behaviour that I think will be essential in future.

I was very fortunate to find myself next to Chris Pissarides at the conference, and at the dinner Ned offered a toast to Pissarides and to Dale Mortensen who was also there, and who had been a young participant at Ned’s seminal conference 40 years ago. Ned then gave a speech deploring the recrudescence of Keyneianism.

On Sat we began with a visit from Fashu Chen and his entourage. It was very interesting being introduced to him and he spoke movingly (interpreted by a colleague) about how in 20 years he had built up from an 8M2 store to a multi-$bn industrial and investment group. His dream when he started was to own a small apartment and a small car and be able to travel, so when he achieved these he started thinking about how he could use his wealth to benefit others, esp from his remote home-town. He has now given $1.2bn (45% of his wealth) to set up the New Huadu Charity Foundation, and Ned is Dean of the Business School he is founding. This prompts the reflection that, however much skill and hard work and talent one puts in to business, substantial success is always also some element of good fortune. However giving your money away wisely requires pure skill, and is much more admirable.

to be continued...

Thursday, November 18, 2010

Mirdanda, Carmen, ASI and Nature

Frantically busy so v little time to blog. But I should at least record:

  • Watched the first episode of Miranda. It is hilarious, and from the filming I can say that the 2nd episode is, hard though it may seem, even better. Catch it if you can.
  • The Adam Smith Institute has published a short article by me on the Negative NI idea. We'll see if this can be made to fly: long-term unemployment is a massive problem.
  • I managed to record most of Carmen from the Met on Tuesday night so I've been able to re-listen to some of the wonderful performances from Nicole and others. What a treat.
  • Fascinating review article in Nature about schizophrenia. It confirms what my son has always said but which seems to have been denied by some "experts", namely that "sychotic outcomes are associated with growing up in an urbanized area, minority group position, cannabis use and developmental trauma. Relative risks are mostly in the order of 2, although as high as 5 in certain subgroups."

Sunday, November 14, 2010

NY, Ned and Nicole in Carmen

Back from a great few days in New York, much of which is un-bloggable.

As I flew out a charming young lady strolled past into first class: we chatted a bit as we were stuck together in the immigration lines at JFK due to a computer failure, and I realised afterwards that she was Lara Stone. Weds was a beautiful day in NY and I ran in Central Park.

On Thurs I joined Bob Pollack and his colleagues for a celebration of the 10th anniversary of the CSSR at the Riverside Church. This was a joyful occasion with colleagues coming in from far afield, and held in the Tower so we had amazing views of the Hudson at night.

On Friday, lunch with Ned Phelps. In addition to discussing our mainstream work, he likes the Negative NI idea. I've been asked to write this up a bit for a think tank (and possibly the Sunday Times) so this is v encouraging and he's offered to look at my draft.

Then on Fri evening to Carmen at the Met where Nicole was singing Michaela. It is a superb production. Richard Eyre was directing, so the acting is even better than usual. Elīna Garanča makes a wonderful Carmen, with a lovely voice and outstanding vocal control (Nicole says she never sings a wrong note) and conveying the right mixture of sexuality, capriciousness, fatalism and determination. Brandon Jovanovich was very strong and convincing as Don Jose, and John Relyea an imposing Escamillo. The transcendent beauty of Nicole's singing never ceases to amaze me: she is perfectly cast and was also acting superbly: you really felt her courage, faith, naiveté and determination. The way in which she beautifully dominated and filled the Met in her great solo area Je dis, que rien ne m'épouvante in Act III, with the extraordinary ability to sing pianissimo with total emotional conviction and yet fill this massive auditorium (3,995 places, sold out) will stick in my mind form many decades (God willing). Sadly there is no recording of Nicole in this role (yet) but it is being broadcast on Tues at 8pm and it may be possible to get it as a podcast.

Afterwards went backstage to see Nicole and met other members of the cast, including Ed Gardner. Nicole says he's one of the most encouraging conductors she has ever worked with, really bringing out the best in everyone, and that the cast is an exceptionally happy ensemble, which I can well believe in the light of the exceptional quality of the performances.

We then joined some of Nicole's friends, including Alyson Cambridge, who was an undergrad with Nicole: a delightful bunch of people. Nicole and Alyson are both doing Elivra in Berlin next year, and Nicole also hopes to be doing a recital in London in Feb.

Monday, November 08, 2010

The new Miranda - Brilliant!

Last night to the BBC Television Centre as a guest of Miranda to watch her filming her new show. I have to say that, on the evidence of this episode, it is completely brilliant. Despite the inevitable delays, the audience were in stitches, and some of the scenes (esp the funeral) will I think remain in anthologies of the best TV comedy of all time.

One of the delightful things about Miranda is that there is a kind of countrapuntal structure in which an idea used in one scene then comes back a few scenes later: a bit like players entering and re-entering in a farce. It is also so obvious that the cast are really enjoying the show.

The second series begins on Monday 15th BBC2 at 20:30 - catch it if you can or on i-Player.

Saturday, November 06, 2010

Biology, Policy and "Mirror"

Fascinating and amusing talk by Oliver Letwin on Thurs night about his career in politics and the present situation. He made several interesting points, in particular that the Coalition agreement had brought in a structure and coherent process to government decision-making that has been very beneficial. Before ministers would tend to make decisions on the fly and were often deflected by "the machine". Now they have a clear framework within which to operate, and if they need to change things they have to go through due process. This basic professionalism in government is surely a good thing, whatever we think of the party politics.

On the way I had tea with Janie Dee, who will be playing the lead in The King and I in Leicester next month. Great to catch up with her and we must go up and see the show.

Last night went to Cambridge for dinner on High Table at Trinity. Two of my former Professors were there, Alan Baker and JWS Cassells, and it was very good to see them. But I sat next to Greg Winter and John Davison. John is Letwin's father-in-law and was pleased to hear that his talk had gone well. Really interesting discussions with Greg, both about science and policy. He rather likes my "more complex than you think, even when..." principle.

Then to see Daughter in her first play at Cambridge, a new work by Luke Al-Rehani called Mirror. This is a rather haunting and somewhat disturbing three-hander in which Faith, Rape and Pain are trapped together in a place out of this world, and have to interact to understand and resolve their predicament. On the first night it was panned by Varsity so on the second night nobody came except two critics, but it had got much better by then and was given a 4* rating which, although I am biased, I think it deserved. The premise and dialogue were thought-provoking, clever and, as far as possible in a play, original. Daughter was perfectly cast for Faith, the reviewer's said her "serenely pre-Raphaelite Faith inspired exactly the right combination of infuriation and appeal, like the Archbishop of Canterbury’s beard" - he will go far! Even the hostile Varsity reviewer thought she "exuded an appropriate amount of serenity".

Last night tonight, catch it if you can!

PS The reviewer was Toby Parker-Rees.

Thursday, November 04, 2010

Wired for God, and Iain McGilchrist

Last night to Theos lecture by Charles Foster based on his book Wired for God. He skimmed quite interestingly through some of the evidence that drugs can induce experiences similar to some reported by mystics, but pointed out that these kind of experiences were by no means central to religion. He finds naturalistic "explanations" of our religious tendencies unconvincing, and suspects that the reason we have the mental/neurological faculties for belief in God is because this is a true belief.

He is on very weak ground suggesting that there are no evolutionary advantages for religion, but of course that doesn't explain it away at all. There are evolutionary advantages in science - does that make it untrue??

There was a response by Susan Blackmore which was frankly disappointing. Although she has an engaging personality, claiming that "everything in the universe is made of atoms and molecules" is not calculated to establish scientific credibility {we actually know that most of the mass in the universe is Dark Matter and Dark Energy}. She doesn't believe in God, but then she doesn't believe in minds either - that is not as "selves" - the thinks the self is an illusion. Her argument for this is that if we were to dissect the brain we wouldn't find a part of the brain which was the "self". This is such a terrible argument (if we dissected a heart we wouldn't find a heartbeat, or indeed love, but that says nothing about whether heartbeats or love exist) that I am amazed anyone could use it more than once. I tried to get her to understand this, and to reflect on the concept of levels of explanation in biology: not sure if she gets it or not.

The highlight of the evening was to be accosted by Iain McGilchrist whom I had last seen when we were at school together. A brilliant scholar he earned a congratulatory First at Oxford in English and became a prize Fellow of All Souls, wrote a book called Against Criticism and then retrained as a doctor. It transpires that he became a psychiatrist and has now written what by all accounts is a brilliant book called The Master and His Emissary which is about the relationships between the left and right brain. It is great that we are in touch again and I will read it with great interest.