Archive for August, 2018

Relative Momentum Portfolio Rotation to Minimize Bear Market Risk

Friday, August 24th, 2018
  • Relative momentum rotation between strategic assets and cash
  • Relative momentum rotation between a choice of strategic assets and cash
  • Potential problems when using relative momentum approach

One approach to a simple rules-based, mechanical system to manage opportunity and potentially minimize maximum drawdown in Bear markets is to use relative momentum between each strategic asset and a risk-free asset (T-Bills) as an alternative — all-in or all-out of each strategic asset based on its return versus the return of T-Bills.Bear market.

The shorter the evaluation period, the smaller the maximum drawdown exposure — but the greater the trade frequency which is not tax efficient in a regular taxable account, and the more frequent the whipsaw (sort of a head-fake by the market requiring reversal of the position with positive loss or opportunity cost).

The longer the evaluation period, the larger the maximum drawdown exposure — but the lower the trade frequency, which is more tax efficient in regular taxable account, and the less frequent and potentially less costly the whipsaw.

Ways to attempt the avoid the worst of each evaluation period length while attempting to capture some of the best of the each evaluation period may be to use an average of shorter and longer periods, or to step into and out of strategic positions in phases by using more than one evaluation period length.

Relative Performance Rotation With the Swensen Portfolio:

Let’s look at an example, using the Swensen Reference Portolio as a base case with 2018-08-23 data.

Swensen is the CIO of the Yale Endowment, who proposed his “Reference Portfolio” as something from which to depart for a personally suitable portfolio in his book, “Unconventional Successs: a Fundamental Approach to Personal Investment”. He did not recommend momentum rotation in that book.

The Swensen portfolio is 70% OWN, 30% LOAN and 0% RESERVE.

This table shows the Swensen portolio as the default allocation, and to the right of that are four alternative allocations based on whether the strategic risk asset has outperformed risk-free cash over a 3, 6, or 12 months rating period, or the average of those periods. Where the risk asset did not outperform, cash is held in lieu of the asset.

(click images to enlarge)

Pink shaded cells indicate that the strategic risk asset is not held and instead the money for that slot is held in Reserve in risk-free T-Bills.

The tables on the left show the total return of each asset over each evaluation period and the excess return. The excess return is the total return minus the return of the risk-free asset (T-Bills) for the same period.

For equities (OWN), the relative momentum method suggests not owning the emerging markets fund at this time, and only owning the non-US developed markets fund if using the 12-month evaluation period.  Total US stocks and US REITs have positive excess return for all periods and would be held at the full strategic allocation level.

For debt (LOAN), the relative momentum method suggests intermediate-term Treasuries should not be owned if using the 12-month evaluation period.  Short-term TIPS are OK for all the periods.

You can operate portfolios, with any chosen set of risk assets using relative momentum as illustrated here. Such portfolios can be rebalanced at intervals ranging from monthly, quarterly, semi-annually to annually.

The longer rating periods may lag too much, and the shorter periods may be prone to whipsaw losses or opportunity cost. The method is intended to capture the bulk of major up trends and to avoid the bulk of major down trends.

This is not a panacea approach, and may or may not improve returns, but probably can limit maximum drawdowns associated with Bear markets without relying on the variable quality of judgement (assuming as is typically the case that market tops are a process of rolling over; not a steep, sudden drop).

Staged In-and-Out Approach:

Let’s say, you were using a staged in-and-out approach to each strategic asset based on three steps governed by the 3-month, 6-month and 12-month relative returns.

  • VTI would be 30% (all-in)
  • VEA would be 10% (1/3 out)
  • VWO would be 0% (all-out)
  • VNQ would be 15% (all-in)
  • VGIT would be 10% (1/3 out)
  • VTIP would be 15% (all-in)
  • CASH would be 20% (residual not in strategic assets).

Critical Note:

This simple relative momentum investing does not include consideration of the fundamental condition, fundamental prospects for or valuation of the assets used and being rotated. That is a key weakness and risk factor associated with this approach unless the assets are pre-selected based on fundamentals and/or valuation.  Presumably, the strategic assets and their allocation weights would have been chosen for a good long-term fundamental, non-technical reasons.  This method is meant merely to modulate exposure to each strategic asset based on performance relative to a risk-free asset (T-Bills).

Building More Complexity:

The approach can become progressively more complex and granular. For example, a portfolio may have a tactical sleeve that involves less diversified assets, that may be higher opportunity/risk or that are expected to exhibit particularly favorable momentum over shorter periods of time.

Such a sleeve might be for the top momentum sector funds, or top country funds, or top ETFs of any type, or top momentum individual stocks, or the top security from a custom list.

In the example below, we created a tactical sleeve that would invest in the top 2 momentum sectors among the 10 Vanguard US large-cap/mid-cap sector funds, and also invest in the top momentum stock among the 5 FAANG stocks [Facebook (FB), Amazon (AMZN), Apple (AAPL), Netflix (NLFX), Google (GOOG)].

In addition, just to make it yet more complex, the default strategic assets are considered for substitution with a related alternative.  In this case we use these alternatives:

  • default S&P 500 (SPY), alternatives S&P 500 pure growth (RPG) and S&P 500 pure value RPV)
  • default total US stocks (VTI), alternatives S&P 100 (OEF) and Russell 2000 (IWM)
  • default total international stocks (VXUS), alternatives DM markets (VEA) and EM markets (VWO)
  • default US REITs (VNQ), alternative  international real estate (VNQI)
  • default US high dividend (VYM), alternative international high yield (VYMI)
  • default short-term Treasuries (VGSH), alternatives short-term investment-grade corporate bonds (VCSH), ultra-short-term, investment-grade, floating-rate debt (FLOT)
  • default intermediate-term Treasuries (VGIT), alternative intermediate-term corporate investment-grade corporate bonds (VCIT)
  • default long-term Treasuries (VGLT), alternative long-term investment-grade corporate bonds (VCLT)
  • default short-term TIPS (VTIP), alternative long-term TIPS (TIP).

Note: This is not a recommendation for a portfolio, merely an illustration of how complexity can be increased.

It is probably obvious that as the number of portfolio slots increases, and the number of securities considered for each slot increases, the decision process moves from one of paper and pencil to one requiring a database return download source and some coding in an Excel spreadsheet – unless you want to drive yourself a little bit crazy.


Life Stage Asset Allocation Reference Timeline

Monday, August 13th, 2018
  • Consensus Life Stage appropriate asset allocation
  • Historical Life Stage asset allocation returns
  • Forecasted Life Stage asset allocation returns

During this long Bull market, some portfolios have been more aggressive than might be expected over the long-term. Now, we are approaching or may be in the 9th inning or the 11th hour, and are exposed to some extraordinary event risks. It may be appropriate to review what the consensus Life Stage allocation is according to major money management institutions.

By Life Stage, we mean, not age, but how many years before you convert from adding assets to your portfolio to withdrawing assets – the Withdrawal Stage of Life.

While people can enter the Withdrawal Stage at any age, the standard reference model is at age 65 with an estimated 30 years to live (for the portfolio to last). Based on that assumption, institutions have designed what they believe are the most sensible and prudent allocations leading up to and following the commencement of the Withdrawal Stage.

This is the model simplified to 4 asset categories that is near the average of the recommendations of 12 leading institutions (excluding the cash level, which tends to be around 2%-4% in later stages):

You may wonder how those allocation levels have performed historically.

This table shows recommended allocations at 5-year intervals before and after entering the Withdrawal Stage, and how they performed over various cumulative periods through July 2018, and over individual calendar years through 2017, plus forecasts by Vanguard and BlackRock.

The data is for indexes, except for international Dollar hedged government bonds which is based on the Vanguard fund with that objective from 2014-2017, and the VALIC foreign government bonds fund (Dollar hedged opportunistically) from 2008-2013.


Even though the stock market may continue on longer than expected, we think those near or in the Withdrawal Stage should give careful consideration to these “rule of thumb” allocation levels; judge strategic aggressiveness relative to them; and if more aggressive, shift tactically toward them.

We’d like to be more safe than sorry over the next couple of years. As we have discussed before, we are using ultra-short-term, investment grade, floating rate US bonds in lieu of other bonds during this Fed rate hiking cycle; probably until early 2019 (maybe longer for the international bonds portion).

[directly related securities:  VTI, VXUS, BND, BNDX, FLOT, FLRN]

Big Changes With New Communications Services Sector

Wednesday, August 8th, 2018
  • New Communications Services sector 49% Facebook, Alphabet and Netflix.
  • Consumer Discretionary and Information Technology provide key stocks to new sector.
  • Communications Services has lower PEG ratio than Consumer Discretionary and Information Technology which provided most of it constituents.

Next month the Global Industrial Classification System (GICS) will launch a new sector: Communications Services. That sector will include the stocks formerly of the Telecom Sector (notably Verizon and AT&T), but also numerous stocks pulled out of the Information Technology and Consumer Discretionary sectors.

These changes modify the relative appeal of the Info Tech and Consumer Discretionary sectors.

This table shows how the top stocks in the new sector (represented by XLC, a new ETF with a head start on the new sector) come from sister sectors Info Tech (XLK) and Consumer Discretionary (XLY).
(click images to enlarge)

Those are some big names leaving both XLK and XLY, going into XLC. Lots of movement among the so-called FAANG stocks (Facebook, Amazon, Apple, Netflix and Google [now Alphabet]). Facebook, Alphabet, and Netflix will compose 49% of the new sector. Apple remains in Info Tech (XLK) and Amazon remains in Consumer Discretionary (XLY).

We will have to rely on simulated histories going forward to compare each of these three sectors to their historical attributes. One such study provided by the sponsor of these ETFs shows Info Tech and Consumer Discretionary becoming more expensive relative to their simulated 15-year histories; and Communications Services less expensive than its history than the other two.

The PEG ratios (P/E over 3-5 yr forecasted earnings growth) is most attractive for Communication Services at 1.1, versus 2.0 for new Info Tech and 1.5 for new Consumer Discretionary.

This is the full list of members of the new sector along with their weights.

Very recently Communications Services underperformed the other two simulations, because it holds FaceBook, which took a steep dive in late July from which it has not yet recovered.

A backtested plot by Bloomberg of the returns of the new and two recomposed sectors shows each of the three simulations outperformed the S&P 500 since the market bottom in 2009 through April 30, 2018.

This ETF has one of the fastest adoption rates. Even though the sector is not yet official, XLC has raised $368 million in just 2 months of existence. Expense ratio 13 basis points.