June 15, 2022. The ECB called a special meeting to discuss surging bond yields of peripheral member states. With the inflation problem in Europe, you would have thought the ECB would have welcomed higher rates. Not so. The ECB has a very difficult task. It needs to maintain financial and economic stability, keeping a lid on consumer prices while making sure that member states’ interest rates don’t diverge too much. This is very difficult to do. 

One strategy to encourage convergence is to restart unconditional LTROs. The problem child is Italy, whose bond yield spread with bunds began the year at 136 but spiked to 243 last week. The LTRO would allow Italian banks to buy BTPS and finance them via repo with the ECB. The problem with this is that Italian commercial banks are already full up on BTPS. 

Another approach is for the European Stability Mechanism to launch CDOs investing in member countries’ sovereign bonds. The collateral would be managed by independent third party managers with performance fees and risk retention, while the ESM itself would fund first loss and mezz with a fund funded by Euro members according to their capital key. The senior IG bonds would be fair game for private investors and would likely find an audience. 

I’ve been attending ESG conferences these last couple of months, speaking, a bit tentatively, at some of them. I’ve been thinking a lot about ESG and don’t really know what to make of it. It seems to try to capture many things in many ways and in doing so makes things complicated and hard to do. There are too many contradictions, complications, competing needs, and dilemmas. It’s too complicated for me. When I speak to other ESG investors or fund managers these days, I just ask them what they are trying to achieve. If they are trying to make things better somewhere without making a big mess somewhere else, that’s good enough for me. Whether I think they can do it is another matter.

The trade-off between returns and impact has been another major topic that’s increasingly come up. I used to think that ESG would make more money on a risk adjusted basis but that thinking was too theoretical. Hearing the thoughts of other investors, experts and fund managers, I’ve come to tweak my thinking. Or at least consider an alternative view. I do believe that any long term investment must consider all the facts and factors, environmental and social ones included. Why invest with half the picture. In shorter time frames, is it possible to abandon sustainable investing? In 2021 investing in most extractive industries like energy and metals would have made a lot of money but this could not have been foreseen. Once the risks were considered, perhaps it wouldn’t have made sense to abandon environmentally conscious investing. All things considered, is it that any part of an optimal strategy is itself optimal? Mathematics suggests it must. If we buy this argument then investing sustainably must yield better financial results. What about concessionary returns then? I think this is not an investment concept. It is instead the application of business as a tool to a non-financial problem. Whereas a social or environmental problem might be solved non commercially and funded through philanthropy or through the public purse, business presents a more efficient and lower cost solution if for nothing more than from the financial and operational discipline that business brings. So what I used to mash into a single concept is now two, in my mind. Investing commercially requires us to invest sustainably. Efficient philanthropy requires us to apply commercial discipline. To be successful at either, we need the clarity of this distinction.

In the immediate months following the COVID shock in 2Q 2020, Fed and market had expected inflation to be transitory, basically rising from 2% to 4% and fading back to 2%. This proved to be wrong. I had expected a peak of 6-8% fading to 3-4%. From June 2020, we shorted treasuries to hedge all of our duration in our fixed income portfolios, sometimes running net short positions.

I was wrong. I now think inflation will peak early 2023 at 10-12% before fading to 4-6%.

I do not expect a recession but i do expect consumption to weaken. Growth will be supported by investment.

Reorganization of supply chains to make them more robust against accidents, geopolitical events, mercantilist policy, and greener will increase capital expenditure and investment. Increased defence spending will also boost income and output.

The boost to income and output will not be balanced. Income will rise but output will not be available for conventional public consumption. This will cause an oversupply of money for the consumer economy encouraging inflation.

The less efficient and greener supply chains will add to costs and even as they fade, will persist long enough to change inflation expectations. Time is as important as levels.

So, I expect higher inflation but not recession although the nature of growth will change

Data is pretty topical these days. Everybody, including I’m sure, some automata, are talking about big data and AI. I wanted to make a couple of very general observations.

If you have a lot of data but they all look the same, or are pretty close to each other, then the information the data contain is very little. In statistics, we require that when more and more data are added to the dataset, the variability of the data as a whole must grow, not diminish. Otherwise the explanatory or predictive power of our model is weakened. It’s a technical point but one that is illuminating when the data fail to meet the standard. Inferences begin to vary wildly as new data is added.

If we have multiple variables or factors in our model, then these factors must be sufficiently independent of one another. If we have 2 factors which are somehow related, then they can be replaced by one factor. If they are not, then the model again behaves very erratically.

The art in the science is discerning how diverse the data are while remaining relevant. Diverse but irrelevant data are as useless as concentrated and correlated data.

This very narrow phenomenon in statistics can be generalized. If you observe someone only when they are hungry you might conclude that they are  a glutton.

Ukraine needs to invade Poland at once. And subsequently surrender. This will make Ukraine a part of Poland and thus NATO.