Snapshot of Weekly Allocation & Performance (SWAP): 14 Oct-17
Saturday, October 14, 2017
Strategic allocations (17.4%):
Long tech/Nasdaq via AAPL (initiated 11 Oct on iPhone X optimism? and positive seasonality)
Basket of EM equity longs: China small caps, Brazil, South Africa (initiated 11 Oct)
Basket of EM equity shorts: Egypt (initiated 13 Oct on limited room for more positive sentiment), India (reduced 12 Oct post-good/benign CPI)
Basket of FX longs: EUR, MXN (initiated – 11 Oct on NAFTA-related cheapening), ZAR, JPY, CHF
Basket of FX shorts: CZK-EUR, HUF-EUR, GBP, RUB, CAD
Tactical allocations (54.7%):
Basket of EM equity longs (Argentina, Pakistan, Qatar) and shorts (Turkey)
Short Spain-Catalonia: Ibex mini-futures
Long USD-KRW (new trade – initiated 13 Oct on modestly attractive valuation 1129 and some risk of communicated DPRK missiles around start of China Party Congress)
SGD hedge (26.8%) via short SGD-CNH (realized – 10 Oct, re-engaged – 13 Oct post-MAS seminannual meeting on stable-to-stronger RMB during Party Congress and typically while USD is appreciating), short SGD
Week of 14 Oct: +3.4% (prior week +6.0%)
Positive contributions from:
Long Gold (realized – 13 Oct, looking to re-engage)
Long Brent oil (realized – 13 Oct, looking to re-engage)
Long China small caps, Brazil, Qatar equities
Long Argentina equities
Short India equities (realized – 12 Oct)
Short Spain Ibex
Based on my Conviction-Correlation model (will explain this in a future 'Strategist's Toolkit' blog post), I should have: Turned even more long Qatar, engaged in long Silver and Copper. The USD short was signaled and implemented.
ZAR: Although I closed a short ZAR-JPY trade, I should have gone further and went long ZAR outright.
Turkey: I could have also expressed a tactical short USD view by over-riding the model to go tactically long TRY and XU100 (Borsa Istanbul).
Gold and oil: These trades were profitable but most gains were accrued earlier in the week. I should have taken profit sooner, then re-engaged on Thu at cheaper levels.
Rank of Conviction-Correlation model output (Predicted) vs Actual week-on-week returns:
Learnings/What struck me:
Short USD view proved profitable, but I should have engaged more in long EM FX, especially ZAR and – tactically – even in TRY (despite the strategic short view).
GBP volatility during the week signals that the investment community is highly polarized over views/market implications here -> Stay away or play tactically into upcoming EU Summit.
US PPI on Thu was a mini-preview of market reaction to US CPI on Fri even if outcomes were different (stronger core PPI -> stronger USD, weaker core CPI -> weaker USD).
USD weakness post-very small CPI downside surprise -> Either the market is grasping for reasons to short USD, long UST duration or – some observers say – ‘CPI is the new NFP’ given Fed internal doubts over its models (so far no sign as yet that Fed is discounting their models). Note however that many G10 FX tested, but failed to break resistance after knee-jerk rallies. We’ll need to see a new catalyst for momentum to continue these next few weeks.
77% market pricing-in for a December Fed hike is too high, too distant given data-dependence (namely CPI and tax cuts).
NZD was a stealth outperformer late in the week in sign of overly bearish positioning post-inconclusive election results.
It was worth it being awake 2am late Wed night (SG time) for FOMC minutes – not only to gauge market reaction/interpretation, but even just for this: