RBC’s head of cross asset strategy Charlie McElligott dons his full wonksh equity analyst hat today, exposing “The April Effect,” and the specific risk of “nasty double whammy” of levered long-short unwinds should momentum stall.
Quick note of warning as we transition into the new quarter, with potential for major thematic / sector / factor reversals in stocks. The following observation regarding April seeing a seasonal ‘momentum’ factor market-neutral strategy unwind is ‘equities wonk-ish,’ but with real potential cross-asset impact. As we know, the Fed watches equities because there are potential implications with regards to broad US ‘financial conditions’ on consumer and economic confidence. Thus, the scale of potential equities volatility does matter across macro, especially in light of general buyside portfolio ‘crowding’ / ‘high beta’ exposure (into ‘growth’ right now especially) which could exacerbate the dynamic.
COMMENTARY: I have been monitoring a phenomenon seen in recent years (notably of escalating ‘violence’) where ‘momentum’ factor market-neutral strategies experience tremendous volatility / drawdown. Huh? Basically this is saying that “the stuff that’s been working” over the prior trail 12 months—both long and short—experiences a significant unwind. The Dow Jones Momentum M/N index has been -1.8% on average over the past five years and down -2.6% on average over the trailing 15 years…while the three year trailing average in the month of April is an absolutely brutal -4.5%.
Why does this unwind occur? For sure there are idiosyncratic macro factors at play at times throughout this window (i.e. you can probably ‘throw out’ the -27.2% strategy return in April ’09 as markets violently reversed off the crisis lows, or the recovery in crude post Yellen’s ‘weak USD policy dovish pivot’ last year swinging-us from deflation favorite ‘anti-beta’ to cyclical beta ‘value’ during this same period)…but it would seem that there is a ‘signal’ being generated in the market (perhaps it’s an expression of rotating into laggards ahead of “sell in May”?) which is driving a quantitative strategy factor rotation with strong seasonality into the quarter-change.
So if this phenomenon were to occur again, what are the implications? We’d need to watch both the ‘high flyers’ and ‘biggest losers’ over this 12 month period for reversals. On a generic GICS sector level, that would mean that Financials (+30.4% over trailing 1 year period), Tech (+23.2%), Materials +17.1%), Industrials (+16.0%), Consumer Discretionary (+11.6%) and Energy (+11.1%) could be exposed from the long-side (for underperformance risk), while Healthcare, Utes, Consumer Staples, REITS and Telcos would be the sectors to watch from the short side (potential outperformance). Obviously from the long-side, the 12m window captures a lot of the run which ‘value’ saw last year (cyclicals, inflation)…so with regards to current portfolio allocation, ‘growth’ would then seemingly be most dangerously positioned for a drawdown, especially ‘secular growers’ (MS pointing out that Tech has accounted for ~50% of the S&P’s return YTD, with those 5 FAANG stocks at 30% of the index return!).
Why is this EXTRA critical into this upcoming April? Well, “factor exposures”…that’s why. AlphaBetaWorks and their recent analysis of 4Q16 hedge fund 13F’s shows us that “…nearly 70% of the hedge fund industry’s long equity risk comes from factor crowding.” And of all of those potential factor inputs, the data shows us that 50% of HF relative factor variance is your ‘Beta’ or ‘Market’ exposure…a.k.a. hedge funds’ ‘long’ equity portfolio characteristics are of exceedingly ‘high beta’ right now. As such, “HF Aggregate thus partially behaves like a leveraged market ETF, outperforming during bullish regimes and underperforming during bearish ones.”
Right-o. So basically, there is risk of a nasty “double-whammy,” because IF this momentum long-short were to unwind with violence, not only would popular longs get hit and popular shorts rip higher….but because the ‘longs’ are ‘high beta’ and the ‘shorts’ are ‘low beta,’ you’d get that ugly ‘leveraged’ market impact as well!
DOW JONES ‘MOMENTUM’ MARKET NEUTRAL SEASONALITY SHOWS AN UGLY APRIL ‘UNWIND’ TREND: -1.8% on average over the trailing 5 years, -2.6% over the trailing 15 years and most-acutely, -4.5% over the past 3 years.
‘HIGH BETA’ (‘MARKET’ FACTOR) IS THE MOST CROWDED HF FACTOR: i.e. Equity HF’S are running very ‘high beta’ long portfolios. As such, a potential factor reversal in ‘momentum’ could see a painful ‘shakeout’ of this crowding.
Below chart and table from AlphaBetaWorks:
And now a final thought – perhaps the macro data divergence between ‘animal spirits’ soft data (beating at a 2 z-score rate) vs relatively ‘benign’ hard data (avg of housing & real estate, industrial sector, labor mkt, personal household sector and retail & wholesale ‘surprise’ indices z-scores essentially running ‘slight beat’) will ‘true-up’ come April, which could then act as the ‘macro trigger’ to exacerbate this dynamic further?
‘HARD-‘ VS ‘SOFT-‘ DATA MEAN-REVERSION IMMINENT? 2 Z-SCORE AVG BEAT IS TOP OF 10 YEAR RANGE: Potential catalyst? Trump “animal spirits” fade in coming months with no further clarity on tax policy until Summer ‘deadline.’