Investing Amid Low Expected Returns Ch. 5: Defensive Factor Blending Without Becoming a Performance Chaser

"Every factor sleeve begins as a diversifier and ends, unless deliberately restrained, as a performance-chasing machine." — the practical failure mode of factor investing

Investment Background

Chapters 1 through 4 established the problem and refused the easy exits. Everything is expensive; adding a seventh expensive asset does not help; reaching for yield is destructive; and enduring the drought is the price of admission. This chapter takes on the one remaining legitimate portfolio-side response, and it must be handled with unusual care — because it is the response most easily corrupted into exactly the behaviour the book warns against.

The legitimate version is this. Alongside the market beta that dominates any sensible portfolio, there exist a small number of return sources whose payoff is driven by something other than the discount rate: value (cheap assets versus expensive ones), quality and defensive (profitable, stable, low-volatility companies), momentum and trend (positions that follow price direction across assets), and carry (holding higher-yielding assets against lower-yielding ones). Ilmanen's argument for them in a low-return world is not that they are magic. It is that their returns are structurally unrelated to whether the discount rate rises or falls, which makes them one of the few honest diversifiers of the correlated-expensiveness problem from Chapter 3.

The corrupt version is what most investors actually do. They read that factors work, look at which factor has performed best over the last five years, buy it, hold it through the beginning of its inevitable drought, and then rotate to whichever factor performed best over the new trailing five years. This is not factor investing. It is performance chasing with better vocabulary, and it reliably underperforms the plain index it was supposed to improve upon.

The Wall Street Translation

Why Defensive Factors Specifically

For a retiree, the factor menu should be filtered by a criterion most factor literature ignores: what does this do in the specific scenario that ends a retirement?

Factor	Return source	Retiree relevance	Honest drawback
Quality / Defensive	Profitable, stable, low-leverage firms are systematically underpriced by investors seeking lottery-like payoffs	High — shallower drawdowns, and drawdown depth is what sequence risk feeds on	Lags badly in speculative bull markets
Value	Cheap assets have a higher starting yield, the Chapter 1 mechanism applied cross-sectionally	Moderate-high — directly addresses correlated expensiveness	Droughts exceeding a decade have occurred
Trend / Managed futures	Positions follow price across many assets, producing a long-volatility payoff profile	Moderate — has historically performed in extended drawdowns, when it matters most	High fees, complexity, and long flat stretches
Momentum (cross-sectional)	Winners persist over intermediate horizons	Low — high turnover, tax-inefficient, crash-prone at reversals	Poor fit for a taxable retiree
Carry	Compensation for holding higher-yielding assets	Low — correlated with the same crisis risk the retiree is trying to hedge	Fails in exactly the scenario it is held for

The filter produces a short list: quality/defensive first, value second, trend a distant third for those who can tolerate its costs. Everything else adds turnover, fees, and tax friction in exchange for a diversification benefit a retiree cannot reliably use.

The reason quality/defensive ranks first is worth stating plainly. It is the only factor whose primary benefit is drawdown reduction rather than return enhancement, and drawdown reduction is what a retiree actually needs. A retiree does not need to beat the index. A retiree needs to not be forced to sell 30% below the peak in year six of retirement.

The Discipline That Separates Blending from Chasing

Ilmanen's practical guidance amounts to a set of structural commitments made before the sleeve is funded. Every one of them exists to remove a future decision.

• Blend, do not select. Holding a diversified mix of defensive, value, and trend simultaneously — rather than picking the one that seems best — accepts that you cannot predict which will work. The blend's whole purpose is that its components take turns.
• Size it small and fix the size. A sleeve of 10–20% of the portfolio is large enough to matter and small enough that a decade-long drought is survivable. The size must be written down at inception, because any size chosen later will be chosen in reaction to recent performance.
• Rebalance into weakness mechanically. The blend only works if you top up the component that has been losing. That is the entire mechanism, and it is the one thing discretion will refuse to do.
• Set an evidence standard, not a patience limit. The exit condition must be a change in the factor's mechanism or an unacceptable fee, never a run of bad performance. Chapter 4's rule applied here.
• Prefer cheap, transparent, low-turnover implementations. The gross premium is modest. A 90-basis-point fund with 120% turnover consumes it entirely, and the investor ends up bearing tracking error for no compensation.

The Fee and Complexity Arithmetic That Kills Most Sleeves

This deserves its own accounting, because it is where the majority of retail factor investing quietly fails.

The honest long-run gross premium for a well-constructed defensive or value tilt is modest — a small number of percentage points at most, and considerably less in most periods. Now subtract: the fund fee, the trading costs of turnover, the tax drag on realized gains in a taxable account, and the behavioural cost of the investor's own mistimed entries and exits.

In a large fraction of real retail implementations, the sum of those subtractions exceeds the gross premium. The investor has taken on tracking error, complexity, and the psychological burden of watching a sleeve underperform the index they read about daily — and received nothing for it.

This arithmetic is why the chapter's conclusion is conservative. A factor sleeve is worth holding only when it is cheap, low-turnover, diversified across factors, held in a tax-advantaged account where possible, and small enough that its drought does not threaten the plan. If any of those conditions fails, the plain low-cost index core is the better instrument, and there is no shame in that conclusion.

Execution Rules

1. Write the sleeve's size, composition, and rebalancing schedule before funding a single dollar, and never revise it in response to performance. A sleeve sized at 15% of the portfolio, split across defensive and value, rebalanced annually, is a plan. A sleeve whose size changes because "value has been working" is performance chasing wearing a policy statement.

2. Screen every candidate fund on cost and turnover before you look at its returns. Reverse the normal order deliberately. If the total expense ratio plus estimated turnover cost exceeds roughly half the factor's realistic gross premium, reject the fund regardless of how attractive its track record looks — the track record is largely the premium you will not receive.

3. Blend at least two uncorrelated factors and rebalance between them annually. Hold defensive and value together, and if using trend, hold it as the smallest slice. Then top up whichever has fallen. The rebalancing between factors is where most of the sleeve's realized benefit comes from, and it requires no forecast whatsoever.

4. Place the sleeve in tax-advantaged accounts wherever possible. Factor strategies generate turnover, and turnover in a taxable account converts a modest premium into a modest premium minus taxes, which is frequently zero. Asset location is one of the few free improvements available.

5. Never let the sleeve exceed 20% of the portfolio, and never fund it from the safe assets. The core remains a broad, cheap, globally diversified index plus the cash and government bond sleeves that exist to survive a crash. The factor sleeve is a modest hedge that sits alongside that core — it is not a replacement for it, it is not permitted to consume the crash-protection assets, and a retiree whose plan only works if the factor sleeve delivers does not have a plan.

Retirement Application

The honest role of a factor sleeve in a retirement portfolio is narrower than the marketing suggests, and understanding that narrowness is what makes it safe to hold.

It is not there to raise the return of the plan. A retiree who is counting on the sleeve to close a funding gap has built a plan that depends on a modest, unreliable, decade-variable premium arriving on schedule. It will not arrive on schedule.

It is there to change the shape of the outcome distribution. Specifically, a defensive tilt aims to produce shallower drawdowns, and shallower drawdowns interact directly with the mechanism that actually kills retirement plans. Sequence-of-returns risk is not driven by average return; it is driven by the depth of the drawdown that coincides with the early withdrawal years. An allocation that gives up some upside in speculative rallies in exchange for losing less in the crash is trading exactly the right way for a retiree.

That is a real benefit and a modest one. Set against it are real costs: tracking error that must be endured while friends discuss index returns, added complexity in an account that a surviving spouse may one day have to manage alone, and the ever-present risk that the sleeve becomes the thing the retiree fiddles with.

The final honest position: a small, cheap, blended, mechanically rebalanced defensive sleeve alongside a low-cost index core is a reasonable choice for a retiree who genuinely will not touch it. For a retiree who will, the plain index core alone is better — and it is better by a wide margin. The correct answer here depends more on self-knowledge than on the factor evidence.

Risk Management

• Chasing risk: The dominant failure mode. It is prevented only by a written, size-fixed, blended sleeve with mechanical rebalancing, never by intention.
• Fee erosion: The premium is modest enough that cost is not a secondary consideration; it is often the deciding one. Audit total cost annually and be willing to exit a fund purely on fees.
• Drought risk: Factor droughts exceeding a decade have occurred and will occur again. Size the sleeve so that this outcome is disappointing rather than damaging, which in practice means the plan must fully survive the sleeve returning nothing.
• Complexity risk: Every additional holding is a decision point in a future crisis and a burden on whoever inherits or eventually manages the account. Simplicity has real, underweighted value in a retirement portfolio.
• Crowding risk: A premium widely known and heavily invested may be compressed relative to its historical size. Plan on receiving less than the backtest, and treat any implementation that requires the historical premium in full as unsound.

What This Chapter Cannot Do

This chapter cannot tell you whether factor premia will persist at their historical magnitude. Reasonable, well-informed people disagree, the evidence is genuinely contested, and the honest expectation is compression rather than disappearance. Any allocation whose success requires the full historical premium is built on the least defensible assumption in the chapter.

Nor can it identify which specific fund will implement a factor well. Construction details — how a factor is defined, how often it trades, how it handles sector concentration — vary enormously between products carrying identical labels, and those details frequently matter more than the factor choice itself.

What it can do is set the boundary conditions under which holding a factor sleeve is defensible rather than harmful: small, cheap, blended, mechanical, tax-located, and permanently subordinate to the index core.

---

Key Takeaway: Defensive and value tilts are one of the few honest responses to correlated expensiveness, because their returns are not driven by the discount rate — but the sleeve degenerates into performance chasing unless its size, composition, and rebalancing rule are fixed in writing before it is funded. Its job is shallower drawdowns, not higher returns, because drawdown depth is what sequence risk feeds on. Keep it under 20%, keep it cheap, fund it from equities rather than from the crash-protection assets, and accept that it is a modest hedge beside the low-cost index core and never a substitute for it.