Friday, 22 August 2014

Janet Jackson Hole

09:56 Posted by The Thalesians (@thalesians) No comments

I am probably not the first person to note that firstly, Janet Yellen and Janet Jackson share a first name and secondly, that Janet (Yellen) is speaking at Jackson Hole. Hence, the juxtaposition of the three words Janet, Jackson and Hole, provides a suitably confusing headline to this article. The one way to clear up this misunderstanding is obviously to place a comma somewhere in the statement. Hence, the correct headline should read "Janet, Jackson Hole".

Traders are actually trying to do the same! Should they take a view on what is essentially a binary event (Janet Yellen speaking at Jackson Hole)? Will Janet (Yellen) be hawkish or dovish (see the plot above which gives relative number of hawkish versus dovish news references for Yellen)? Or should they place a pause or a "comma" in their trading and simply cut their risk, taking no view over the event?

The question does more broadly open up the question of how traders should deal with scheduled event risks. From a systematic trading perspective, if a longer term trading model has been backtested over a large history which includes such events, the bias might be to simply let a model run. Alternatively, we can investigate what has happened over these specific events historically to create an overlay model for a specific event risk. For example, I wrote a Thalesians quant paper yesterday on this (cited below) and found that generally stocks have rallied over the past decade of Jackson Hole events. However, whilst there might be similar events in the past, notably previous Jackson Hole speeches by a Fed chair, the expectations around each event can be different. Our sample size is also quite small, which complicates such analysis.

We can see this behaviour of how varied expectations can be, in all scheduled events. Notably this can be observed in the way in which implied volatility event add-ons can vary. For example, before payrolls, EUR/USD overnight implied vols could be marked up by as much as 5 vol points, if the market is particularly jittery about the number. Other times, it could be much lower. Unfortunately, vol tends to be overpriced over scheduled events. Hence, the notion of always hedging scheduled event risks through options can be very expensive over time.

A more realistic answer is actually to combine ideas of cutting risk and using options to manage risk over events, using some discretion in when to use either or both. In addition, we can often have cases where we might want to actively take risk over binary events if we have a particularly strong view. The key of course is to be flexible, rather than having a very rigid approach when it comes to scheduled events. Whilst events might well be scheduled, the outcomes can be far from "scheduled".

If you liked this have a look at Trading Thalesians, my book on investing (mixed in with a bit of ancient history).

Thalesians - Jackson Hole in one - Historical market moves during Jackson Hole - 21 Aug 2014 (available for Thalesians clients only - sign up for free to shortened versions of my papers)


Post a Comment