Macro Portfolio Strategy: Adjusting our gamma

Much like a theme needs orchestration & arrangement to be fully realised in music, so to does it in our portfolios….

Up until recently, we hadn’t done much on macro portfolio strategy.  We’ve decided to open the black box a fraction, starting with September 29th article discussing how our risk indicator affects our decisions, as well as the back half being on the composition of our hedges.

From the article:

Our preference for sometime has been to periodically rebalance a basket of save-haven assets weighted on cheapness.  Shorting Australian & Canadian dollars present favourable hedges to our portfolios, and we’ve recently implemented this beta-adjusted as our largest gamma-long positions.  Next in queue is Gundlach’s Total Return fund, and finally a smattering of gold, Build America Bonds and long-dated Treasury bonds.

To go to the core of our thinking, we are not really trying to square off the delta in our portfolio simply with negatively correlated assets.  Rather, we are trying to buy gamma sensitive assets when they are cheap.  The distinction in wording is subtle, but realised in a portfolio, entirely different.

Let’s analyse these gamma-sensitive assets — ways to find some exposure to safety:

• US treasury notes & bonds
• US dollars
• Gold
• Options
• Value:Growth equity spreads

(Take a look at the value:growth spread (IWD:IWO) – does that look like the long-bond, or what?)

So, the question that we have to ask ourselves:  which assets will move the most in response to safety-seeking market behaviour?  Specifically:  we want to buy gamma, not just take symmetric hedges.

Behind this is the concept of gamma dispersion — we measure the degree which gamma-sensitive assets are correlated.  To do this, we first simply divide each asset by its long-term trend to find which assets are above or below it, as well as relative to one another.  This essentially de-trending the movement to find the oscillations.

Are you surprised at how disparate the gamma-sensitive assets at any given time?  They frequently disagree with one another — sometimes in direction, but other times in degree.

The extent to which they agree is measured by finding calculating the average de-trended oscillation values, and then sum the variances.

This produces a graph which tells us how synchronised gamma-sensitive assets are:

Can you say:  mean reverting?

Finally, we also rather like holding assets that have underlying bullish cases with a positive carry – government bonds, for instance.  Buying gamma protection that way as opposed to put-buying is generally far more effective:  you actually stand a chance of winning if you needed the protection or not.

So, right now, we can see that gamma protection of all kind is very expensive at present.  Only gold provides us much in the way of cheap hedge.

As a consequence, we’re less willing to take on gross exposure.  Additionally, we’re shifting assets out of dollars, which we’ve done very well on but are no longer cheap, and into gold & BABs.  Gold is probably the most dubious gamma-sensitive asset.  It’s use as a safety bid is fairly new, and the spot price over marginal production cost represents around a 35% premium.  Consequently, we’re more guarded in loading up on gold.  On the flip-side, you can buy BBN (Build America Bond CEF) at a 10% discount to NAV (52-week average is 3.71%) which is yielding 7.81% (29% leverage), and we are encouraged by the thesis put forward by Guillermo Roditi here and here.  It’s average duration is 15y.  So, backing out the leverage, that is a 292bp spread to 20y USTs – which is actually 30bps more than what they actually yield.