Kevin Khang

About Me

I'm a financial economist and I currently serve as Head of Active and Alternatives Research with Vanguard's Investment Strategy Group. My current research areas include active portfolio management, portfolio construction, direct indexing, tax-aware investing, and household finance. My work has been published in the Financial Analysts Journal, the Journal Of Investment Management, the Journal of Portfolio Management, the Journal of Retirement, the Journal of Beta Investment Strategies, and the IMF Economic Review. 

I received a PhD in finance from Northwestern University, studying under Jonathan Parker and Annette Vissing-Jorgensen, an MA in economics from the University of British Columbia, and a BAH in economics from Queen's University at Kingston. Prior to my current role, I held portfolio construction, risk management, retirement research, and consulting roles with Vanguard, BlackRock, SSGA, and other financial services institutions. 

Publications and Working Papers

New & Coming Soon

Decade-long lower-for-longer rate environment + desire for higher returns + inflationary regime (and Fed hikes) = current environment where sound risk management of negative convexity becomes front and center in active muni investing. 

For direct indexing investors, the first years' volatility environment holds the key to the overall tax loss harvesting experience. Whether the first years overlap with a bear market determines the total loss harvest potential and the loss harvest contour over the long-term. Those who prize consistency in loss harvest may achieve it with additional cash contributions. 

Active Management & Portfolio Construction

Once realistic transaction costs and investment constraints are accounted for, no alternative optimization strategy consistently dominates the simple 1/N allocation in factor investing. Successful factor timing is required in order for optimization-based strategy to outperform 1/N. 


Factor premia could have been doubled if investors were rebalancing their factor portfolios daily, instead of monthly, during market cycle transitions. Knowing when to rebalance aggressively is a prized skill for factor investors.


Anchored to either slow-moving or fast-moving time series, industry-standard risk models tend to generate inaccurate forecasts when volatility regime changes. By real-time monitoring cross-sectional dispersion of forecast accuracy of diverse model specifications, it is possible to engage in timely and disciplined transitions between slow-moving and fast-moving models. 


Industry-standard risk model's output is a point forecast of the future volatility with embedded uncertainty around it. This uncertainty is about 20 to 30 % of the forecast itself. This ratio is useful for informed risk-taking, especially in times of volatility regime change. 


We investigate how the balance sheets within the financial sector adjusted during the 2008 crisis, as mortgage- and asset-backed securities (MBS and ABS) shifted around the system and unconventional monetary policies were introduced. Hedge funds delevered significantly, while commercial banks levered up to absorb the MBS and ABS.


Tax-aware Investing

Tax alpha is as diverse as there are diverse investor profiles, ranging from effectively zero to well above 300 bps, and is quite predictable. This calls for a thoughtful customization on 1) how much to allocate to tax-loss harvesting (TLH) strategy  and 2) whether to harvest losses with direct-indexing.


Effectiveness of a TLH strategy varies significantly over two margins: 1) breadth of loss harvesting universe (from sector funds to individual securities); and 2) frequency of harvesting (from annual to daily). Daily scanning for harvesting with individual securities reaches maximal TLH effectiveness in all volatility environments.


We examine how to use direct indexing (DI) by answering two common questions confronting prospective investors. Question 1. How do investors create space for DI in their portfolios, and how much customization can they pursue with DI? Question 2. What is the optimal tax-loss harvesting scanning frequency for capital-gains-rich DI investors? 


Retirement Research & Household Finance

Examining American Community Survey migration records, we show that generations of retirees may have tapped into housing wealth by relocating to a cheaper housing market, extracting about $100,000 to shore up retirement funding. We estimate that one in every four retirees may be able to benefit from using this strategy over the next 10 years.


Sustainable withdrawal rates can vary considerably, depending on the levels of desired bequests and portfolio depletion risk, and asset allocation, highlighting the need for customization rather than relying on a single rule of thumb. 


We perform a scenario analysis on sustainable withdrawal rate (SWR) for investors in today's environment: 2.8% if stock-bond correlation turns positive, inflation and bond market volatility rises, and inflation remains elevated; and 3.3% in the best case scenario. All other SWRs require a departure from the consensus return outlook.  


We quantify the impact of "sequence-of-return" risk--the risk of receiving a concentrated series of poor returns--on newly retired investors. They are 31% more likely to outlive their wealth, have 11% lower retirement income streams, and/or leave 31% smaller bequests.  


During the housing boom of the 2000s, owner-occupants' decision to make future housing investments--remodeling and investment property purchase--was driven by the strength of appreciation on their primary residence. Consistent with the investors extrapolating the housing boom into the future, this effect was stronger during the frenzy (mid-2000s) and more pronounced for younger investors. 


Word-of-mouth effect contributed to the formation of the mid-2000 housing bubble. Existing residents of a community were 24% more likely to purchase investment properties once an experienced real estate investor moved into the community; this effect was particularly strong during the 2002-2006 period and especially in the "sand" states (Arizona, California, Florida, and Nevada) where the housing bubble was most pronounced.