Usually we like to lead off our free newsletter with a trading idea exploitable with our product range. And this time we were intending to explore a forex trade based on this Bloomberg piece by Narayana Kocherlakota, former president of the Federal Reserve Bank of Minneapolis.
But, frankly, negative rates leave us with more questions than answers when looking at trade design. How heavily does it weigh in any given currency pair? How should a trade horizon be estimated? Do market participants, as Mr. Kocherlakota says, really only need to "understand and fully support" what central banks are doing by going negative?
Might be worth noting also that negative rates, obviously, represent punishment of savers. Anyone looking to buy annuities in the next 3 to 5 years is going to be short-changed unjustly.
Additionally, it is not clear even banks will go along with this. Here is a report that Commerzbank is considering simply hoarding the cash - rather than lending it - to get away from ECB negative rates. That is truly saying something about the state of either European banking or Commerzbank.
So at this time we prefer to cogitate some more and restrict our reasoned pair suggestion to what's in the next premium newsletter.
In the meantime, we'd welcome feedback on forex potential trades centered on negative interest rates!
We release the next issue on 20 June at 10h30 GMT and include:
A focus on Global ETFs with a great pairs candidate. Includes a tailored strategy file and complete narrative analysis
A review of one of the most pragmatic pairs trading research papers we have seen for a long time. We actually were obliged to call it a "gem".
Useful tools for analysing news sentiment
Subscribe here before the 20th to get it - we include a 30-day free trial.
What a time to be long-only!
With equity markets near all-time highs it looks like directional trading is the place to be. For now. But if your gift for timing is fallible consider expanding your hedging strategies using ArbMaker products.
For that, this month we have a pretty special offer:
on ANY permanent license, including bundles.
Expires 30 June 2016
(Apply the code by clicking the "Got a code?" text during checkout)
And we've already applied the discount to our single license subscription range. Also good until month-end.
Taken with a 30-day free trial to our spanking new premium newsletter that provides guidance, real-time analytics, flash alerts and downloadable portfolio/strategy files it's a great way to go about affordably designing hedges against directional risk.
The authors make the key points that cointegration alone does not guarantee profitability, nor does it necessarily identify economic substitutes (considered the true targets of a relative value trade). They also find that money neutral pairs are really where the profits are to be found.
As a bonus the authors also give this quote:
“We find that despite being commonly regarded as superior to the Engle-Granger method, the Johansen method yields poorer trading results”
It's tempting to point out at this stage that ArbMaker structures money neutral trades via its time-varying beta functionality. And it does so using Engle-Granger (E-G) methodology.
However, notwithstanding the quote above, opinion is strikingly divided across the body of academic research when it comes to E-G vs Johansen. We're not immune to an implementation of the Johansen method as a complement to E-G in the future.
Our own pragmatic opinion is that what counts is interpretation of the data and the marginality (or preferably not) of the cointegration readings - whichever methodology is being used.
Anorak's corner: out of sample testing
Conventional wisdom suggests that weak out-of-sample results mean unreliable in-sample results. Yet there are two cases in the context of ArbMaker in which in-sample data is not always biased to detect spurious predictability.
The first involves loss of information and less predictive power in small samples. In samples of small size, loss of information may result in an out-of-sample test falsely rejecting valid in-sample results.
The second relates to reconciling instances where the split period used for the training turns out not to be a period over which the pair is cointegrated - whereas the whole period is.
Overall the message might be this: traders could usefully take the view that both in-sample and out-of-sample test results be viewed as informing the pursuit of tradable pairs rather than one being the underwriter of the other.