South Korea is essentially an American colony since the Korean war.  I think Trump administration’s policy toward North Korea is allowing North-South talks to proceed without American involvement and essentially the American attack plans on North Korea will backfire and Korean unification process will begin with quiet influence of China in the back.  This can be avoided, but I don’t think Trump’s people have the competence or the mindset to do much but rush the ‘South Korean independence’ from America.


Hedge funds are sharply focused enterprises but they have a gigantic Achilles’ heel.  They promote a culture of secrecy and competition rather than one of collaboration and nurture.  In this respect, technology companies fare better.  And for this reason, there are projects that would be profitable for which technology companies are superior to hedge funds.  Ambitious projects with good profit opportunities with nurture and collaboration are much more likely to produce good outcomes and success in tech companies than in hedge funds.

I am a slow thinker who only realised in the past few years that stochastic volatility is more or less a universal feature of not only economic time series but probably for all ‘socially constructed’ time series.  Others had not been so slow.  I just found some stream of research that solves the optimal portfolio problem assuming a Heston SV model for dynamics of the risky asset.  OptimalPortfoliosHestonSV is a representative paper and I won’t try to sort out the field because to know that such problems are solved was good enough for me.  There is a disconnect between practical money managers and these sorts of theory problems for good reason–often academics are too busy with doing fine tuning and do not pay attention to the issues of the importance and clarity of the fundamental usefulness of the issues to non-academics.  The key point to recognise clearly is that every time series is not just SV but SV with jumps in price and volatility.  A clear solution to the optimal portfolio problem with these assumptions therefore should be STAPLE for portfolio managers because this is a solved problem.

This is obvious but worth mentioning for die hard Hayekians.  The Fed employed intervention by Quantitative Easing.  Hayek’s argument regarding how central planners operate on too little information given the signaling mechanism of prices is interesting but wrong if you look at the laws governing macroeconomy — to whatever extent laws can be claimed to have been found.  They worked well enough to have a recovered economy.  This was Keynes’ genius, not socialism or communism.  So Keynes will live on and perhaps Hayek and Von Mises will have secondary roles still.

The will to dominate is not leadership but a degenerate lack of civilization. There is something Neanderthal about the basic need for some people to dominate what they cannot understand and therefore fear for that reason. I operate best around civilized people who necessarily are respectful and not tainted by a diseased psyche filled with the raw will to dominate assuming that such a will makes them great human beings. They are dangerous bullies who have the effect of stamping out creativity and lucid genius when we are lucky sometimes to achieve hard-won lucidity, for believe me, I would kill today to have the level of lucidity in mathematics that I was born to have. Will to dominate has a negative influence on self-lucidity as well as lucidity of those one associates with.

Stochastic volatility models are our best models for financial time series dynamics and there has been enormous advance in our understanding of how to estimate SV model parameters even without the option chain price data.  Therefore they can be thought of as necessary part of any signal dataset for algorithmic strategies that have probably not yet saturated the algo trading strategy space.  In the late 1990s there was a great deal of development in methods of estimating SV parameters of which one of the prominent ones was MCMC Bayesian approach.  The following paper provides some guideline:Bayesian-Analysis-SV.  Gregor Kastner has the R package ‘stochvol’ that implements an MCMC inference algorithm for a basic SV model.

I am Shiva, the penniless ruler, and I shall show you the destroyer of worlds, the world of physics where Oppenheimer said, I am Shiva, the destroyer of worlds before dropping bombs on Hiroshima and Nagasaki, but the worlds that tempt destruction most is a scientific revolution in physics, raising Eternity from Oblivion and perhaps Justice from her incarceration.