Explaining Today's Market Action (Hint: Blame Risk Parity)

From Charlie McElligott, head of cross-asset strategy  of RBC


“You have to pick your poison -- more risky assets or a more balanced basket with high leverage…There’s a philosophical
underpinning on why risk party has worked and why it should continue to work…Every time people talk about it as a leverage bond
portfolio, I just cry….It’s not leveraged bonds.  It’s a leveraged portfolio. ”

--Edward Qjan of PanAgora Asset, who “coined” the term “Risk-Parity”

 

Risk-parity 101: Leverage historically “low volatility” asset classes (e.g. fixed income) / subsectors (utilities) alongside historically “higher volatility” assets (stocks, EMFX, or the tech sector) to better “balance” your multi-asset portfolio’s risk-allocation (which in a 60/40 equity/bond portfolio would see 90% of “risk” concentrated on the equities side).  Net / net, the strategy uses leverage to allocate “risk” instead of allocating assets for diversification.  Different parts of the economic cycle see different underlying asset class allocations (i.e. the current ‘low growth, low inflation’ backdrop), and now you’re cooking with grease—an “all weather” portfolio, ahem. 

Me, last month in “RBC Big Picture:”

“The issue is when the historical vol is based around a 35 yr bond bull mkt....bastardized by QE and NIRP....that when you get taper tantrums, crowded position unwinds or a rogue inflation print, that hyper-leveraged long in fixed income goes pear shaped as rates go ‘wrong way.’  The other dynamic with ‘risk allocation’ based on volatility modeling in this current environment is angle is that via CB policy, nearly all cross-asset vol has been crushed--so per the model, these fund types “get longer” the assets now too on trailing vols which have been at extreme lows—even “traditionally risky” assets like equities.”

Thus, being extremely-qualitative and making overly-simplistic generalizations on RP (across this wide universe of strategies somewhere in the AUM universe of $1.2T via applying 3x’s leverage), it still remains intellectually honest / accurate to say that when we see a portfolio with ~70% weighting towards fixed income (short term / long term rates, credit and inflation protected assets) and with said leverage, a sharp upward move in rates is going to cause short-term stress.   And when we see a concurrent move lower in risk-parity trading vehicles like S&P e-minis (Spooz -20 handles from earlier highs--check), Gold (a 3.5SD selloff in the ‘inflation proxy’ favorite, see below--check), EM (EEM and EMB as global growth reads—check) and US long bond (2SD move over the past 3 sessions) it becomes a simple & observable fact-pattern.  Many of these funds are still very long any asset class where trailing vols have remain suppressed—ironically, bc there is so much demand for ‘carry strategies’ in said environment, even the recent “Taper Tantrum 3.0” was viewed as a vol event to be sold into for additional carry, and thus quickly muted.

And as a reminder—it’s not just RP necessarily, but various other “vol / risk control” products (Morningstar had “risk control” AUM in Q4 ’15 of ~$750B…unleveraged).  As highlighted in the past, “risk control” products are extremely popular post GFC period with institutional market participants with long-term investment horizons and long-standing liabilities—think pensions (defined benefit plans) and insurance companies (variable annuities).  These products are pitched as offering a “smoother path” of asset returns. 

(Quick explanation of how these things work: these products typically have two components: 1) an underlying index to track and 2) cash.  When vol increases, the allocation moves out of the underlying index and into cash (sliding scale, dynamic based off of present vol target).  When vol decreases, the allocation moves more weight into underlying index, and less in cash // If the vol of the underlying falls below tgt levels, the exposure can then be leveraged to capture “extra upside”…and vice versa.  For a typical index, the benchmarks are set at 5 /10 / 15% risk targets (conservative, moderate, aggressive)…thus, in the case of the 15% risk control model, the max leverage is 150% (ratio= tgt vol level / historical vol).  When historical vol falls to 12%, the index can allocate 125% (15%/12%) into the underlying index etc.  End of the day, these products offer lower max drawdown, lower max return, lower vol, lower beta to S&P—you can see why they’re so easy to sell and keep growing AUM.)

In the case of today, it is not as simple as “VIX” or implied vol as the “trigger”…esp as VIX itself is pretty “sleepy”  at just +4.5%.  The trick is that there is also a very large component of the ‘risk-control’ product universe that actually uses “market beta” as the trigger…and that is how both risk-parity and risk-control strategies tie-into today’s single stock trade—via my whipping-boy YTD—the crowded “low vol” / equity “bond proxy” trade.  The “risk control” products which use beta as their trigger end up kind of looking like “anti-beta” factor market neutral, with slight nuance.  Instead of being “long low beta / short high beta” (as you’d see in the mkt neutral” strat—of course, with heavy leverage—and which are -2.2% over the past 5 days alone, which is really really bad for a mkt neutral strategy), these products swing their allocations of a diverse basket of stocks based upon beta to act as their “gearing” (without using outright leverage—which is a selling point to retail investors).  Thus, in high volatility, they go shift overweight “low beta” stocks like….telcos, REITs, staples and utes.  In low vol, they go overweight ‘high beta.’  So on a day like today when ‘low vol’ and ‘low beta’ are eviscerated, but on a ‘meh’ VIX and a down-stock market, these types of products don’t work.

As referenced earlier today in the “RBC Big Picture” note, I highlight how the 30day historical vols in both USMV and SPLV were both HIGHER than that of SPY.  Well, we see this same insanity playing-out today as the “low vol” trade is further unwound today:
 
DEFENSIVE SECTORS = SPANKED ON RISK-PARITY / RISK-CONTROL UNWIND:

LOOK AT THE MONGO UNDERPERFORMANCE OF “BOND PROXY” YIELD PLAYS TODAY:

FROM AN ASSET-CLASS ‘RISK THERMOMETER’—SEEING RISK-AVERSION ASSETS UNDERPERFORM IN A RISK-OFF TAPE:

GOLD SEES 3.5 STANDARD DEVIATION SELLOFF EVENT PER RETURNS BACK TO 2013:

 

The punchline is that the passive / smart beta / risk-parity / risk-control systematic universe often times ARE the entities in the market causing counter-intuitive trading behavior, such as today’s price-action.

It doesn’t mean the strategies don’t work—as evidenced by the fantastic YTD performance of nearly the entire RP universe YTD for example (thank you, CB Gods).  But what it does mean is that when their enormous AUM and leverage are mechanically and unemotionally “triggered” to de-leverage, you get said “bull in China shop” heavy-handed allocations (this stuff nowadays makes me long for the “Good Harbor” dynamic allocation days of old!). 

Here is to the simpler-times that are loooooooooooooong gone.