Look, the thing you need to understand about Citi is that the bank’s pool of strategists is divided into two camps.
There are the credit folks who live in the real world, and then there are the equity folks who quite literally live in “FANTASY” land.
I capitalized “fantasy” there not for emphasis but because that’s the acronym Tobias Levkovich proudly coined in a note out earlier this month and profiled here.
Headed into Q3, the bank has updated their ETF factor recos and if you’re familiar with the above-mentioned “FANTASY” note, it probably won’t come as a complete surprise to you that they are still Overweight Momentum.
Basically, Citi thinks staying long/Overweight Momentum smart beta products is the way to go.
Of course the thing about a strategy based on momentum is that there’s a certain circular logic to it – it’s a bit tautological.
And besides that, one of the biggest concerns in markets right now is the extent to which growth, momentum, and low vol. have become synonymous with the stocks that are driving benchmarks to new highs. Again, there’s something tautological there, and indeed that’s precisely the problem. Because what it suggests is that if/when things get going in the “wrong” direction, all of this will be dumped mechanically.
That contention is backed up by this, from Bloomberg’s Dani Burger:
Visualization of today's NDX selloff and momentum-y unwind. The more a share gained over the past 12 months, the more it's selling off today pic.twitter.com/ZFYheAEh11
— Dani Burger (@daniburgz) June 29, 2017
someone's been reading SocGen's Andrew Lapthorne.
— Heisenberg Report (@heisenbergrpt) June 29, 2017
Speaking of Andrew, here’s what he had to say about this very same problem earlier this month in the wake of the June 9 tech selloff (the one Goldman definitely didn’t cause just like your dog definitely didn’t destroy your living room while you were at work):
The sell-offs themselves are not particularly unusual, but the uniformity of the prices moves all on the same day indicates a market driven by price chasing momentum, with investors heading for the door all at the same time. Indeed, those S&P 500 stocks which sold-off on Friday were almost all from the strongest performing decile over the previous 12 months (the r-squared on the S&P 500 line in the chart below is 85%). Within Nasdaq the relationship is even stronger at 95%.
Such a uniform sell-off strikes us as systematic, especially as the relationship weakens once you look at the broader and less liquid Nasdaq composite. For price chasing investors, Friday’s plunge serves as a warning; when it’s time to head for the door, you better move fast.
Well anyway, Citi still thinks Momentum is the place to be headed into H2, and below you can find some of the color from their latest ETF perspectives piece…
Mind the Momentum
With our ongoing overweight allocation to Momentum, this quarter’s Factor Focus provides a more in depth look at the underlying characteristics and performance dynamics of Momentum smart beta products. The key takeaway is that Momentum needs to be differentiated from the action in QQQ (Nasdaq 100).
First, as we’ve noted in previous work, the “FAANG” stocks (Facebook, Amazon, Apple, Netflix and Alphabet [Google]) have been a major contributor to the S&P 500’s YTD return. Yet, their inclusion in several of the lead “Momentum” ETFs is limited. In some cases, such as with MTUM, this is a function of recent rebalances as described later in this section.
Expanding to the “FANTASY” stocks (coined by Citi US Equity Strategist Tobias Levkovich, Facebook, Amazon, NVIDIA, Tesla, Alphabet, Salesforce and Altaba [Yahoo]), doesn’t change the Momentum exposure that much. At issue is our underlying ETF mantra: “know your index”. Index definitions and rebalance activity will influence which stocks are characterized as Momentum. In turn, fundamentals and investor sentiment at the single stock level will influence factor ETF action over shorter timeframes.
The SPDR Russell 1000 Momentum ETF, ONEO, which “tilts” towards stocks with higher Momentum exposure in the Russell 1000, includes most of these stocks. But in aggregate, carries less exposure than the S&P 500.
Second, another implication of varying methodologies, parent indices and rebalancing schedules is differentiated sector exposure (see Figure 3). Our decision to remain overweight Momentum (MTUM in our model) is heavily influenced by the increased weight in Financials, which occurred with its recent rebalance.
Trends in sector allocations have been somewhat similar since ’15 across funds. In general, weights to Consumer Discretionary and Staples have declined while exposures to Financials, Industrials and Tech have climbed (see Figure 4).
Unlike most Momentum factor products, FDMO applies a sector neutral weighting constraint. The only meaningful historical allocation change occurred when Real Estate was broken out of Financials and treated as a separate sector.
Third, looking at the five Momentum factor products highlighted above, MTUM has posted the most sizable outperformance relative to the S&P 500 year-to-date, returning +18.3% vs. SPY’s +9.0%. PDP, FDMO and SPMO have outperformed as well, but by smaller margins, gaining +12.0%, +10.2% and +9.9% respectively. So far, only ONEO has underperformed (see Figure 5).
Figure 6 provides a breakdown of MTUM’s strong year-to-date relative performance against the S&P 500. While differences in sector weightings have positively contributed to excess returns, the main driver of outperformance has been underlying stock screening and weighting variations.
Figure 7 details sector performance and performance attribution details of MTUM and the S&P 500. First, Energy, which has carried a modest weight in MTUM prior to the MSCI Momentum Index rebalance, has exerted minimal drag on portfolio performance. In the S&P 500, Energy has hindered overall index returns by more than -1%. Second, Tech, a heavy weight in MTUM, has provided the most noticeable boost to absolute and relative returns. Third, MTUM sector returns are leading the S&P 500 in 10 of 11 categories. Consumer Staples is the lone outlier.
Fourth, going beyond traditional sector and industry group breakdowns, Figure 8 examines MTUM and S&P 500 year end allocations by constituent forward returns. To do this, we first sort 31 Dec ’16 S&P 500 constituents by year-to-date performance and then review their allocations in the S&P 500 and MTUM. Our findings show that MTUM is heavily overweight top year-to-date performers while having minimal exposure to the bottom 40% of index performers (see Figure 8). While sector weightings are partially responsible, Figure 6 and Figure 7 above show that stock screening has played an even more important role year-to-date.
The key takeaway is that investors need to be cognizant that even for a “technical” factor, like “Momentum”, index definitions incorporated into ETFs will vary. Understanding index definitions, and related macro/micro influences, on ETF composition is critical to effective use of single factor fundamental ETFs, in our view.