authored by

July 2021

This commentary is in two parts.  To begin, I will outline the guiding principles of the Great O’ Neill’s risk management philosophy - Critical State, Second Order Chaos, and Interest Rates. I will follow this by going into a little more detail to demonstrate how the Great O’Neill quantifies and manages risk.

This commentary is in two parts.  To begin, I will outline the guiding principles of the Great O’ Neill’s risk management philosophy - Critical State, Second Order Chaos, and Interest Rates. I will follow this by going into a little more detail to demonstrate how the Great O’Neill quantifies and manages risk.

In the simplest non jargon language possible the Great O’Neill views market risk as follow:

Markets DO spike up or down violently (critical state).  Risk is UNAVOIDABLE and cannot be calculated mathematically (second order chaos).  Risk is probably under priced (negative interest rates).

Critical State

Aside from acknowledging that it is hard to manage risk I have only considered the emotional (Greed and Fear) inputs.  I think that The Great O’Neill has the right mix of greed and fear. This informs our technical approach to markets.  From a mechanical point of view, I consider all markets to be in a critical state or on a knife edge.  It is like a penalty shoot-out at the knockout stage in a soccer match.  The next penalty kick will decide a major result – win or lose.  It is unpredictable and yet massively important.  It is probably not a good idea to live your life as though you may be run over by a bus or win the national lottery at any moment.  But it is a good state of mind from which to manage your financial risk.

I found Mark Buchanan’s book ‘UBIQUITY –WHY CATASTROPHES HAPPEN’ very useful.  He took a complex subject from physics (critical state) and explained it in layman terms.  He explains that in markets, as in nature, extreme events happened all throughout history.  We wrongly assume that the natural state of markets is calm.  In truth, markets like nature are in a critical state.

We can know for certain that market crashes and earthquakes will happen in the future.  We cannot know when they will happen.  The natural state of markets is critical!

Second Order Chaotic System

Weather forecasters opine on the weather.  Their opinion on whether it rains tomorrow or not has zero impact on the actual outcome.  What will be will be, irrespective of a forecaster’s opinion.  In other words, the weather does not care about opinions.  Weather is a first order chaotic system.  In a first order chaotic system what we think will happen has no effect on what will actually happen.

Markets care greatly about forecasts and opinions.  Capital markets are a second order chaotic system.  The specific valuation level of oil is partly determined by market participants opinion on the valuation.  Those opinions are often based on factors beyond supply and demand.  Many hedge fund managers that focus on trends and momentum will express a bullish or bearish view on oil prices for no other reason than the price per barrel of oil has been going up or down for a number of days / weeks / months.  This may encourage oil producers to increase or decrease supply.  Equally oil consumers may reduce or increase their consumption based on hedge fund driven price action.  We can see that the number of factors affecting oil prices can compound very quickly.

Even if it were possible to calculate the myriad of individual factors affecting the demand side and supply side of oil, we would still only have achieved half the work.  Our second task is to try to calculate the reactions in the market-place to all these factors.  This creates endless feedback loops.  If we knew that tomorrow OPEC would announce a cut in supply, we could not guarantee that oil prices would go up.  Even if we could calculate the probability of an event like an OPEC supply cut, we cannot determine whether it will push prices up or down.  The future remains unknowable, measuring the likelihood of futures events, (e.g. an OPEC supply cut), and measuring the impact on price of that event is impossible.

I think it is productive to acknowledge that we cannot know how risky an individual trade is.

Interest Rates

I have written more extensively on this subject in three commentaries (Interest Rates Parts I, II and III) and I believe they are well worth reading.  For the purposes of this commentary, I am focusing on the effect that artificial, or government mandated interest rates have on risk metrics.

The ‘risk free rate of return’ (“ROR”) is an integral component of most financial models and calculations.  What exactly the ROR is and how it is measured is subject to some arcane debates amongst economists.  However, the simplest explanation is probably the best.  It is the benefit of doing nothing with your money, or put another way the opportunity cost of doing something with your money. Negative interest rates and inflation literally means that there is a cost to doing nothing with your money. Doing nothing means you lose money.

Negative interest rates increase the supply of capital available to invest in risky assets.  In other words, the demand for risk seeking capital is being met with an ever-greater supply.  This in turn lowers the price paid / received at any given level of risk.  For example, lowly rated junk bonds trade at lower prices (yields) today because of negative interest rates.  This creates a feed-back loop leading the market to think that risk is lower, when in fact it is not.

In the Great O’Neill’s view central banks have pushed interest rates lower, ergo risk is trading at below fair value in today’s markets.  I do not know what the fair of value of risk is.  However, my working assumption is that it should be higher.


 Our guiding principles lead the Great O’Neill to view nominal trade position size as the primary determinant of risk.  We do not consider some products to be less risky than others.   In other words, US Government Bonds futures are not considered to be less risky than oil futures.  This approach sets us apart from most other hedge funds.   In my view it is a fallacy to collate historical volatility data in order to form a view on future risk.  Markets have continually surprised participants with unexpected spikes or plunges. Financial markets are in a ‘critical state’.  No market or product is less immune from an unexpected price spike or plunge.

There are some obvious immediate liquidity constraints in some markets over others. Illiquid markets with little open interest are not treated the same as liquid markets with higher levels of open interest.  This is an important secondary consideration for the Great O’Neill.

The maximum amount of leverage used is 1.5 times the nominal value of the assets under management (AUM).  This exceptionally low leverage rule sets us apart from most of the industry.  The Great O’Neill does impose position limits for correlated market sectors.   The aim is to limit exposure to one product to circa 20% of our maximum book size (AUM *1.5).  Additionally, we aim to limit exposures to highly correlated markets to less than 25% of our maximum book size (AUM * 1.5).

Given below is a typical picture of our nominal position sizes as a percentage of our book.  It is a snapshot from March 31, 2021.   It is our first and crudest estimate of the firm’s risk exposure.




We are short about $144,000 of Soybeans(ZS,14%), that is the nominal size of our Soybean position is 14% of the Assets Under Management (AUM).  We are long WestTexas Intermediate Oil (CL 2.45%).  This is a call option and the nominal is quite low. We do not write naked call options; naked puts are treated as (delta adjusted) futures positions for the purposes of risk management.  Our positions are as follows:

Short Soybeans (ZS), Corn (ZC), Silver (SI), Nifty Index (Indian Stock Market), Coffee (KC), and Cotton (CT).  Long Cattle (LE), Oil (CL), and S&P 500 (ES).   Overall, we are operating at 60% of book.  We have never reached our max exposure of AUM * 1.5.

Our second measure of risk is a little more refined than the first.  The chart below is a simple pivot chart of our exposures.  It shows the losses likely to occur if the underlying market for each position moves adversely by 20% in one day.  This chart does not take into account any positive correlations.  We can see from below that a crash in the S&P 500 causes loss of 2.95% and a gain of 20% in the Nifty Index (India) will cause a loss of 2.4%.


If everything goes horribly wrong, including the American stock market dropping 20% while the Indian stock market shoots up 20% the Great O’Neill will lose 17.2%. Our worst single day since inception is -5.4%.

The next stage of my risk analysis is based on the Dollar.  I ask the question, what if everything drops or rises by 20% tomorrow versus the Dollar.  This question is relevant because many markets are highly correlated today.  This is often referred to as the risk on, risk off trade. This analysis involves more charts which I will spare the reader. The Great O’Neill also takes into consideration the overall risk exposure it has at a given time and potential market correlations when assessing risk.  Judging ‘risk on / risk off’ effects and market correlations is very discretionary. The weekly or even monthly correlations between Soybeans and Corn or S&P 500 and the Nifty cannot be relied upon.  A hard drop in the S&P 500 would probably be mirrored by a drop in the Nifty. However, I think it is not smart to become overly dependent or presumptuous of historical or ‘natural’ correlations.

Most estimates of risk assume a likely or probable range of outcomes.  These assumptions are somewhat limited by human imagination and the use of Gaussian distributions.  How many oil traders at the beginning of 2020 factored minus (negative) $35.00 per barrel of oil into their probable range of outcomes for 2020? Admitting that we cannot adequately calculate risk, forces the Great O’Neill to control risk via reduced leverage and option purchases.

The two charts given above demonstrate our‘stress test’ approach to measuring our exposure and thus risk in the market.  It is not meant to be perfect.  We have not solved the riddle of measuring the unknown.  Nevertheless, our approach to risk management sets us apart from our peers in the SG CTA Index and this may be at least partly borne out in the performance chart below.



JOHN KIERANS – 6 July 2021

return to commentaries