Define market (in)efficiency: the Euro

Euro/USD prices: An anomalous move? Source: thinkorswim by TDAmeritrade

The Efficient Market Hypothesis is a well-known, well-respected theory.  It’s frequently cited by folks in the economic & financial space to justify very conventional, buy-and-hold investment products.  Particularly in the stronger forms of EMH, all publicly available information is immediately reflected in asset prices.

I’ve written before about the impact of ‘known unknowns’ versus ‘unknown unknowns’ in financial markets.  The recent crash in the Euro strikes me as an interesting case study:

• Why is the Euro crashing?  In a term, policy divergence.  The European Central Bank is providing quantitative easing to the Euro-area, which means (all other things equal) higher money supply and a cheaper currency.  In contrast, the Federal Reserve is gradually pulling back from their quantitative easing (i.e. letting their purchased bonds run-off), and thinking about raising interest rates this summer.  Both these actions should mean a stronger US Dollar.
• Why such a prolonged slide?  This is where I think the markets are interesting.  In an informationally-efficient world, I would have expected a big move at the onset of European QE, with not much happening thereafter (perhaps some oscillation around a new equilibrium level).  Instead, we’re being treated to a managed futures manager’s dream scenario: a fairly steady downward trend, without many pull-backs.
• Perhaps behaviour economics can answer?  An explanation could be found in the inefficiencies considered by behavioural economics.  Among the possibilities:
  • Discrete decision points: perhaps each business/investor considers the QE announcement at different intervals, and thus make moves in sequence.
  • Loss aversion/disposition effect: folks long Euros have been holding on, while losses pile up.  Over time, they accept their painful losses at different points based upon relative aversion.

In sum: efficient markets shouldn’t really show these kind of smooth trends.  But the trends exist all the same.  Hence it’s worthwhile considering momentum as a viable investment strategy.

Let there be…vol?

Will there be calm, or more party time? Source: Google Images

So this is probably the last ‘serious’ week for financial markets of 2014.  Some thoughts:

1. Is oil done?  The news seems more bent on $40/barrel oil, or at least $50, so another 10-15% down move from here.  I’m sure many recognise that the media is generally way late to the party, so perhaps today’s slight recovery to above $58 is putting in the near-term floor.  My momentum models don’t care about the debate, and are staying well-short.
2. Which is right: VIX or S&P?  Last week’s rise in the VIX, from about $12 to about $19, was an outlier move – similar to what happened last October.  So are we due for an exciting, proper sell-off in the S&P?  Or is this morning’s resilience in the index (up about 1%), combined with VIX selling off (down about 3%), the more relevant fact?  On Friday I reloaded on my old favourite UVXY trade, so I’m clearly hoping the latter.
3. I feel bad for being long grains.  My same momentum models have me long soybeans, which has been a pretty good trade so far.  However I can’t ignore the oversupply, which I hear from family in the Midwest.  Another example of how the biggest enemy to a systematic trading approach is probably manual intervention.

In sum: I’d like a quiet week.  My models would prefer a chaotic week – or at least a continuation of that lovely oil trend.  With the remaining economic news of 2014 released this week, combined with rolls/option expiry, I’m guessing there will still be plenty of action.

Ignore John Bogle???

I was reading this Marketwatch piece this morning, and find the topic quite interesting. There is plenty of opinion out there that equal-weighted indices outperform market cap-weighted indices.  And when you look at RSP versus SPY, you indeed see the result.

Why would you choose RSP over SPY?

1. Why index? The conventional reason folks choose stock indices (especially broad-market indices, such as SPY) is to diversify away specific company risk.  We know that holding equity should pay a return in the long run; we just don’t want to get unlucky choosing a bad equity.  So we invest in everything, looking to average returns.  RSP, by investing equal amounts in each of the S&P 500 constituents, has more diversification than SPY.
2. What about rebalancing? The recent article by Campbell Harvey et al. is instructive here.  It turns out that regular rebalancing increases risk versus keeping a static portfolio.  So perhaps RSP pays more than SPY because it’s taking the extra risk.  Indeed, this is reflected by RSP’s 1.11 beta versus SPY: the former takes roughly 11% more risk than the same $ investment in SPY.  At least for the past year, the return of RSP has been about 1.11x SPY, so I guess the risk/return level is about commensurate.  Anyway, I could just suggest buying more SPY than buying RSP.
3. What about momentum? See the same article.  SPY, like other market cap-weighted indices, implicitly take a momentum approach to the market. Because there is no rebalancing in SPY, the fund will automatically allocate more capital to stocks with higher returns (and thus higher market cap).  Given momentum is a lasting source of return, you’re essentially getting a trading strategy for free in choosing SPY over RSP.

So what to choose? If capital were no issue, I’d probably just buy more SPY than going smaller in RSP.  I like the lower management fees (yes, I do agree with John/Jack Bogle on that point), and appreciate the implied momentum returns of SPY.  If capital were an issue, I’d think of RSP like IWM: a way to achieve higher returns for the capital than SPY, mainly due to overweighting smaller companies.

How would you like your returns skewed?

There have been several times in the past where I’m explaining ‘XYZ strategy’ to someone (hopefully they asked me beforehand), and the concept of skewness comes up.  A couple examples:

• Several (successful) strategies lose far more frequently than they win.  It’s not always like playing the lottery…
• Sometimes ‘the sure thing’ trade, which has made money every day, suddenly blows up.

Thus loops in the concept of skewness – how big are losses relative to gains?  On the lottery side, you’re almost certainly going to lose  USD 1 on a game with a (highly improbable) gain of USD tens of millions.  But other examples abound in financial markets:

• Long-only (just about) anything: this is a negatively skewed strategy.  Most months/years you will have a gain, but some months will be TERRIBLE.  Don’t think about the little correction we just had…think about 2008.  It can take years to recoup the losses from long-only: for example, notice that the NASDAQ is still about 10% below its 2000 peak.
• Venture capital: this is a positively skewed strategy, in its most basic form.  The VC fund manager selects (say) 10 companies at an early stage of development.  Financials don’t really mean much at this stage – they could do anything.  The hope is that, out of 10, there will be 1 big winner and maybe a few small winners.  The others are expected to be written off.  So, one gain outweighs the many.
• Volatility selling: this is a classic negatively skewed strategy.  VERY negatively skewed, epitomising ‘picking up pennies in front of a steam roller’.  After premium selling funds lost about 50-70% in 2008 (or went completely bust), several actually hit high water marks in the past couple years.  So it’s a sustainable, if nerve-wracking, strategy.  By the way, insurance products and market making are roughly the same as option premium writing, in terms of performance characteristics.
• Momentum trading: a classic positively skewed strategy.  Frequently momentum is classed as ‘long volatility’, which it is…kinda.  More long gamma…but anyway.  This is a ‘pain trade’, in that most of the time you’re losing money as markets oscillate back and forth and you’re trading with the trend.  Only occasionally do the big trends come; you can’t really forecast them, and you MUST be in the market when they come.  Otherwise this is a losing strategy.

I leave you with the following track records, harvested from Altegris’s managed futures website.  Interesting place to learn about volatility and momentum offerings.

Classic volatility selling strategy characteristics: nice, steady gains punctuated by large losses. Source: http://www.managedfutures.com

Classic example of a higher-geared momentum fund. Notice how the fund spends most of its time below high water mark; this is broken up with infrequent, large gains. Source: http://www.managedfutures.com

Extra credit: those seeking more technical info on return skewness, and particularly how the time-variance of skewness is a function of strategy design, should look at this wonkish paper.

Let’s do some data mining – Death Cross revisited

It’s all well and good looking for the Death Cross signal to tell when to be in and out of the markets…but is that just an artefact of the data?  I’m curious…

We established last time that being long when the S&P500 was above its 200 DMA, then selling when the price closed below the same DMA, was a somewhat ‘meh’ strategy: returns were about the same as being long the entire period since 1950, but risk/return was a bit better using the strategy.  Using the joy of AmiBroker, plus the plethora of free data on offer from Google Finance etc., we can get some idea of how robust the Death Cross really is.

Here’s what I’ve done:

1. Downloaded all historical data for the investable S&P500 ETF – SPY.  This begins Feb 1993, so has a few ups and downs.
2. Using AmiBroker, I simulated a realistic-ish simple trading strategy:
   1. Be long SPY if (an exponentially-weighted) moving average of the price (MA1) is above another moving average of the price (MA2).  Let’s not assume the shorter moving average needs to be higher than the longer for the time being.  BTW, a moving average of 1 = the current price; thus the Death Cross is equivalent to MA1(1) > MA2(200).
   2. If the first MA crosses below the second MA, exit the market until they reverse.
   3. Assume we check these values at the close of each day, and transact on the next day’s open.  No cheating!
   4. Assume a flat $9 for each trade.  That’s about average these days for US brokers.
3. Vary the lengths of MA1 and MA2 over the following ranges:
   1. MA1: 1 day through 80 days, incrementing by 1 day
   2. MA2: 2 days through 300 days, incrementing by 2 days
4. Record the net profit of each run, and place in large PivotTable.
5. Colour said large PivotTable to create the below piece of abstract art:

The closest I come to art. Green = relatively better idea than Red. Sources: AmiBroker and Google Finance.

Lovely.  Each row represents a different MA1 value; each column a different MA2.  The colour scheme helps us see some interesting patterns:

1. Longer averages = better results.  The Death Cross result is that small blue circle, deep in the lower-value area of the chart.  Why?  By having such a reactive average (i.e. the underlying price), each time the price falls below the 200 DMA there’s a lot of buying and selling.  For example, if you unloaded yesterday due to the DMA being crossed, you might very well be getting back in tomorrow.  That’s 2 transaction fees for not much benefit.
2. ‘Sweet spot’ = around 50 vs 100 DMA.  Over this time period, the darkest green area is around the MA1 = 50 and MA2 = 100 days.  I guess this is because those runs run away from the 2008 crash well: the 50/100 run exits Nov 2007, then re-enters June 2009.
3. Still doesn’t beat returns of long-only.  Even the best run has about a 105% total return (excluding dividends) since 1993.  Buy-and-hold SPY returned over 300% (excluding dividends) in the same period.  But the drawdowns are perhaps more palatable – around 18% Max drawdown for the 50/100 run, versus 55% for long-only.

What are the lessons?

• Trading momentum strategies, like the Death Cross, helps the risk profile of an equity portfolio.  It doesn’t help the return profile, at least since 1993.
• Data mining is sometimes very dangerous – if using data analysis to come up with a really good-looking, overfit strategy.  I think it can be a helpful tool to see how robust a strategy is – as in, what if I’ve completely misspecified the strategy?  In this case, you can be really loose about the 50/100 parameters and still come out better than the Death Cross.