Stronghold EUR update: The intense days of March 2020 to cut risk
Head of Equity Strategy
Summary: In this monthly update on the Stronghold EUR portfolio we go through the intense days in March when liquidity disappeared and the model aggressively cut risky exposure to maintain its risk budget. Stronghold EUR was down 6.5% in March as portfolio had cut all risky exposure in credit and thus did not benefit from the rebound during the last week of the month. Read our recollection of what happened during the month and how we view the portfolio's performance this year.
March 2020 will go down as one of the most dramatic months in financial markets and an acid test of the ETF market. The Stronghold EUR model had cut equity exposure to zero in February and was coming into March being the most defensive since the portfolio's inception in July 2017. Despite the portfolio being defensive markets became highly correlated and liquidity became erratic in some the ETFs that Stronghold EUR was exposed to.
Even though Stronghold EUR leaves the first quarter with a negative quarterly return, it has performed decently when comparing to other tactical asset allocation portfolios with similar risk levels. Comparing to the BlackRock EUR portfolios on Saxo Bank's SaxoSelect offering, the Stronghold return for the first quarter lies between the quarterly return for the BlackRock Defensive and BlackRock Moderate Defensive portfolios.
The Stronghold EUR model started March well as credit and inflation-linked government bonds were bid but then the US Treasury market and inflation-linked bonds began a historic dislocation due to margin calls and collapsing inflation expectations. The model had to reduce risk multiple times during the month starting on 16 March selling all exposure to investment grade corporate bonds and majority of exposure to inflation-linked bonds. The spreads had already widened and liquidity was terrible. The trade-off was between high trading costs hitting the portfolio value or staying within our risk budget. The latter is most important at all times so we had to eat the increased friction and trading costs in ETFs. Two days after on 18 March the portfolio's covered bond exposure was cut to zero as mortgage bonds in Europe went in free fall relative to normal volatility; there were no places to hide as investors were in a 'dash for cash'. Even gold became correlated and acted as a risky asset such as equities. The 18 March was an extraordinary intra-week rebalancing as the estimated portfolio risk was too far away from our target. On 18 March asset classes became so correlated that we had to add 0-1 year Euro government bonds with a 20% maximum weight in order for the model to stay within the risk budget. This was not something we had seen during the financial crisis in 2008 during our simulations.
Finally on 23 March the model reduced some of its 10-year global government bonds exposure and increased exposure in Euro 1-3 year government bonds. Despite the aggressive actions of the model the portfolio was down 6.5% in March compared to minus 3.7% for the benchmark. The reason for this difference is driven two factors. The benchmark is static with quarterly rebalancing so trade costs are minimal. Stronghold EUR had to sacrifice some performance to maintain its strict focus on minimizing risk. The second factor is that Stronghold EUR went ultra-defensive with 80% exposure in Euro government bonds across the 0-1 year and 1-3 year segments and zero equity exposure so when the 18% rebound in global equities came Stronghold EUR did not get any return from that. Meanwhile the benchmark has a 35% exposure to global equities.
Year-to-date Stronghold EUR is down 8.6% compared to 7.1% for the benchmark portfolio. With many economists expecting a bigger hit to the economy from COVID-19 than the 2008 crisis financial markets will continue to be volatile and in the case equities continue to fall Stronghold EUR will benefit relative to the benchmark.
Liquidity vaporized in European ETF market
As March unfolded liquidity disappeared in many European ETFs with market makers widening spreads dramatically. The average bid-ask spread in percentage terms peaked during the month at 3.1% for the ETF tracking inflation-linked government bonds. Rebalancing the portfolio to maintain the model's strict risk target caused higher friction and losses than normal as liquidity conditions were catastrophic. Our clients benefited from our access to European dealers through the request-for-quote systems tapping into non-exchange liquidity and thus our ability to navigate was better than if everything had been done through exchanges.
The selling pressure in corporate credit and covered bonds ETFs forced prices significantly below the net-asset-value (NAV) of the underlying bonds although the pricing in the underlying instruments likely had great uncertainty. The ETF used in the Stronghold EUR for tracking investment grade corporate bonds was trading at one point at -5.7% discount to reported NAV in the fund. Central banks stepped into markets with new quantitative easing programmes which helped corporate bonds to recover in value. Other key policy responses from central banks and fiscal authorities helped risky assets to recover some of the losses by the end of the month. As Stronghold had moved to an almost all-in cash position (we use short government bonds as synthetic cash) the portfolio did not benefit from this rebound in risky assets.
As the current portfolio weights suggest (see table) the portfolio is mostly in cash-like instruments such as Euro government bonds with short duration. The portfolio's risk budget is fully utilized with the current exposures but the portfolio has 18.3% exposure across gold and 7-10 year global government bonds which allow the portfolio to switch this exposure into more risky asset classes when the expected returns become high enough. On a positive note many signs suggest that liquidity has resumed to European ETFs which will make rebalancing easier going forward. While we could have hoped for a better first quarter Stronghold EUR has performed decently when we compare the portfolio against other tactical asset allocation portfolios with similar risk levels.