Some integrated thoughts on various aspects of personal financial investments which may be useful to others.
0. Use a sound framework for making decisions
- "To invest successfully over a lifetime does not require a stratospheric IQ, unusual business insights, or inside information. What's needed is a sound intellectual framework for making decisions…” - Warren Buffet
Simple but useful guidance
Use a sound framework - Have a sound intellectual framework for making decisions - Warren Buffet
Focus on a small set of risk factors - Target a small set of risk factors - El Erian
Be research/evidence based - Use research and rational expectations, not hunches – Fama Jr. (update)
Avoid mistakes - Charles Ellis
Implementation Framework
Be holistic - Consider financial assets, human capital, and insurance in investment decisions
Simplify - For ease of management over an investment lifetime
Write it down/commit - Have an Investment Policy Statement (it can help in both good and bad markets)
Understand limitations - Benefits, limitations, and costs of the Fama-French risk factor approach
1. Personalize
- “Personal preferences, circumstances, and abilities affect portfolio construction in a profound manner.” – David Swensen
Structure financial risk exposure to match financial targets
Set specific targets - Personalize financial targets
Factor in labor income risk - Characteristics, international vs. industry hedge, relative benefits, and longevity risk
Match ability, willingness, need to take risk - risk exposure (factor cap tilt, & liability-matching), odds of success
Insure against disruptions to labor income - and for asset protection
Reflect personal sensitivities and circumstances in investment decisions
Risk of real loss - balance ‘deep’ & ‘shallow’ risk, (inflation vs. downside, LT purchasing power vs. ST declines)
Consider home financing in fixed income allocation decisions - as a negative bond allocation
2. Structure exposure to risks rewarded with premium returns
- “The ideal situation is to come up with a small set (three to five) of distinct (and ideally orthogonal) risk factors than command a risk premium.” - El Erian
Dimensions of risk
Risks that have been rewarded with premium returns
Seminal paper - Common Risk Factors in the Returns on Stocks and Bonds, -(additional references).
Market - global, larger premium than 'expected', , outlier risk (debate)
Value - distress risk (debate), crashes, behavioral, consumption risk, theory & practice, ‘margin of safety’
Size - definition,default risk (table, analysis), microcaps, delisiting bias, less consistent size than value premium
Term - risk and return to increasing term exposure
Default - most premium explained by equity factors, default load in stocks higher than bonds, FF summary
Cross-sectional Momentum? - historically large (except Japan), negative correlation with value, factor loads
Time-series Momentum? - volatility scaling effect, stronger in growth than value portfolios, example
Profitability? - other side to value? definition, less inflation but more protection in financial crises & deflation
Minimum volatility? - simply a combination of other factors, in US and global stocks
Will these premiums persist? How can a strategy still work if everyone knows about it? Its not data mining
Have realistic expectations: Volatility lessons
Risks that have not been rewarded (for individual investors)
Leverage (e.g. hedge funds) - inefficient exposure to risk
Manager risk - high cost of hedge funds, uncompensated private equity manager risk for individual investors
High yields bonds - hasn't added mean-variance efficiency or downside risk protection (factor analysis)
Risk diversifier considerations
Event risk - commodity prices add non-financial event risk hedge (1974-75, other examples), risk of CCFs
Peer-to-peer lending - Credit risk
Other unique risk - REITs don’t add much unique risk (REITs in value funds). Diversification benefit.
Structure exposure to risk
“The next step is to assess the stability of the factors and how they can be best captured through the use of tradable instruments.” – El Erian
Summary view
Structuring exposure to equity risk
Market - Its hard to beat Vanguard for pure market exposure
Value - Use mid or small cap value ETFs (or TM value funds), not large value [Updated Factor Exposure]
Small - Use microcap funds , but expect large tracking error
Momentum - Reduce negative momentum via fund selection , consider adding large cap momentum fund,(example)
Profitability – Some micro & sv funds have positive profit loads,more so than midcap value[RAFI US, intl equity],but be cautious in interpreting profitability coefficients, incremental value of adding profitability is small
Component vs. multifactor funds - Tilt is tilt,evidence shows little difference, consider overall portfolio
Reconstitution considerations - More frequent reconstitutions has not historically added to returns
Rebalancing considerations - RAFI methodology may add rebalancing returns,particularly in international
Structuring exposure to fixed income risk
Default -Not needed, negatively skewed, downside shows up at the wrong time. Use Treasuries
Term - Consider more term with high equity allocations, Duration shifting not necessary. (Swensen’s view)
Inflation protection - nominal bond/TIPS split (liability matching) (views on TIPS, DFA, understanding TIPS)
Structure the combination of equity and fixed income risk
Interest rates - Consider extending fixed income duration if have high value tilt
Diversification - Diversify across risk factors, other considerations, rely more on asset class than time diversification
Tracking error - Reflect tracking error tolerance in allocation, particularly with small cap & value tilted portfolios
Downside risk - Hedge risk of different market environments, returns of TSM, large & small value in bear markets
Black swans and fat tails - Minimize effects of negative black swans, reduce negative tail effects.
Time-varying vs static exposure to risk factors
“Deliberate short-term deviations from long-term policy targets introduce substantial risks to the investment process” - David Swensen
Common approaches to time-varying (smoothing) exposure to risk has a poor performance record
Inverted yield curve – yield spreads are weak predictor of equity returns, with wide spreads after inversions
Moving average approach – tax inefficient, lower returns than small cap & value tilted portfolio (Faber paper).
Focus on investment not speculative returns – From Keynes, Bernstein, Bogle, Buffet
Index vs. active management
"The efficiency of marketable security pricing poses formidable hurdles to investors pursuing active management strategies" – David Swensen
The odds of identifying superior managers (net of all costs) is low
While Swensen offers better guidelines than most on the selection of active managers
Odds of identifying superior managers, ex-ante, is low
Comparative exercise on asset allocation, security selection, marketing timing
Berkshire, Longleaf, Sequoia, Tilson, PIMCO Global Multi-Asset, Primecap, Aquamarine, Pabrai, Thaler, Vanguard Wellington, Vanguard Wellesly
Have measured expectations
When risks show up
Don’t be surprised - performance during the ‘Great Depression” (outlier considerations), crash of 1929 (video)
Equity correlations rise - risk premium correlations can turn higher (1930-57)
History - equities, sovereign debt, inflation, ’deep risk’
3. Diversify globally
- “…the global market portfolio is a good place for you to start..” – William Sharpe
A view from Swensen’s – Pioneering Portfolio Management
International diversification works (in the long run)
US equity markets
Historical returns across global markets
US multinationals in international allocation
Non-US developed equity markets
The value effect is a global phenomenon
Consider costs of international diversification
Try to get at least some intl. value and small cap exposure
International small cap options
Consider country rebalancing returns
Emerging equity markets
Risk, not GDP, is priced into EM stocks
Does EM stock dilution matter
EM stocks have higher expected return than developed mkt stocks
But, EM stocks are also negatively skewed (A 20 year perspective)
Keep fixed income in US Treasuries to reduce EM’s negative skewness impact on an overall portfolio
Frontier equity markets
Frontier market considerations
No diversification benefit beyond EM, so far
Non-US Fixed Income
No need for foreign bonds, they provide unwanted currency fluctuations, added costs, lower protection (a view from Swensen and Bernstein)
4. Minimize costs
- "...realize the highest possible portion of the return earned in the financial asset class in which you invest
– recognizing and accepting that that portion will be less than 100%" - John Bogle
Taxes
Be tax sensitive in investment decisions (the anatomy of tax efficiency)
Compare tax efficiency across funds (large and small, midcaps)
Avoid the tax cost risk of active management
Consider ETF turnover in tax-efficiency assessments
Assess stand alone vs. share class ETF tax efficiency
Expenses
Use low expense, high quality fund companies
Limit active management (which adds on average a 0.67% annual cost over passive alternatives)
Style drift
Use R^2 and alpha as selection criteria to minimize style drift
Use more midcap than small cap value to reduce sampling error risk and migration impacts
Don’t use fixed income (credit) ETFs due to rebalancing risk (NAV vs. price drift)
ETF risk not linked to issuer risk (eg. Barclays or State Street)
Reconstitution arbitrage, trade costs
Do DFA funds have lower trading costs?
Does DFA have a cost advantage? (excl. advisor fee) [estimate 1 and 2]
Does DFA add alpha?
5. Policy targets and sample portfolios
- "Over the long haul, what matters is factor exposure and expense" - Bill Bernstein
Investing for Retirement
* The above are the same as my policy targets
since developing an Investment Policy Statement (IPS)
at start of 2003. Reasons for the 0.2/0.4 size/value load target, other considerations, performance of value tilts during bear markets
Note: The factor loads used to construct the ETF and DFA Tax-managed portfolios listed below are estimated, while those for the DFA core/vector and “RAFI” portfolio comparison are assumed (not estimated) and therefore less precise (due to lack of available data). The annual performance of the first two portfolios is expected to be closer to each other than with the second two (as in the recent historical back-tests). Differences will likely be due to: (i) differences in factor exposure across international markets, (ii) differences in term and default exposure in fixed income markets, and (iii) the less precise determination of the portfolio factor loads for the second two portfolios. However, if the factor loads used (in above link) represent the long-term characteristics of the funds, then the long-term returns across all the portfolios are expected to be similar. The back-test returns are derived from actual fund returns when available, and the respective underlying index returns prior to fund inceptions.
Sample portfolios to try to match above policy targets
Portfolio sample with iShares, Vanguard, Bridgeway funds [ETF Portfolio]
Portfolio sample with DFA Tax-Managed funds [DFA Tax-Managed Component Portfolio]
Sample portfolios aligned to policy targets with close to zero negative momentum load
Portfolio sample with RAFI, Vanguard, iShares [RAFI/CRSP/MSCI Portfolio]
More detail on the factor load estimates
Likely performance dispersion of the sample ETF portfolio and actual performance
- Sample portfolio with slightly different factor exposure but similar long-term expected return
Portfolio sample with DFA core/vector funds [DFA Core/Vector Portfolio]
Portfolio sample with RAFI funds [RAFI]
The RAFI data is from Arnott’s Fundamental Index book, supplemented with data from Powershares. Estimates of US RAFI value and size loads are from earlier presentations by Arnott, which were similar to Bernsteins', the same are assumed for Intl markets. Here's a more recent factor load matching of a RAFI portfolio and a DFA vector portfolio, estimated over a shorter time period
Investing for College Education
Utah currently provides the best 529 option for those who want to use a globally diversified value tilted portfolio for investing for college education. Here is a sample portfolio that matches closely the policy targets (reflected above) of global equity diversification (50:37:13 for US:Non-US Developed:EM ); and a small cap and value tilt (0.3/0.4 size/value load) which are fairly constant across the age-based allocations. The only thing that varies is the the stock:fixed income allocation, and the composition of fixed income with a declining term exposure (increasing inflation protection) over time. This component of fixed income approximately matches the estimated first withdrawal share by the time it is due. While half of equity are in Non-US currencies, all fixed income is in US dollars, and the US dollar composition of the portfolio increased from 55% (0-3yrs) to 90% (+19yrs) when dollar payments are need.
Utah 529 Customized Age Based Allocation
6. Ensure consistent implementation
- "The discipline to stay the course with an asset allocation is in all likelihood the greatest determinant of returns in the long run, more so than asset allocation itself" – Larry Swedroe
"Many aspects of investing are fun, but your future wealth isn’t a game. You should manage it in the most cold-blooded fashion. Emotion, pride, ego, dreams, and nightmares have nothing to do with the process, although some investors rely on little else" – Peter Bernstein
The challenge, importance, and discipline of staying the course
There will be inevitable periods of underperformance
Use allocation bands to rebalance back to policy targets (consider valuations -sample approach backtest)
Tax loss harvest
Keep emotions out
Survival strategies
When considering changing funds
Measure performance – XIRR vs Modified Deitz
Use goal-based performance monitoring:
- Performance relative to value path to long-term targets,
...which is dependent on meeting:
- Savings targets (annual $ amount)
- Returns targets (personal long-term policy target = 7.5% annualized)
- Factor exposure and expense
Common Risk Factors in the Returns on Stocks and Bonds
Value and Momentum Everywhere
The Other Side of Value: The Gross Profitability Premium
Index and Enhanced Index Funds
The Riddle of Performance Attribution: Who's In Charge Here—Asset Allocation or Cost?
Useful reminders
Ten of the Best from Bogle
Ten of the Best from Bernstein
Bogle videos
International, small-cap and value premium
Fund returns, costs
Dollar cost averaging
Active managers
Financial webcasts/interviews
Fama on why small caps and value stocks outperform
David Swensen – on equity orientation and diversification for institutional investors and individual investors, conversation
Historical performance
Dimensional Matrix Book 2020
Credit Suisse Global Investment Returns Yearbook 2009 - Keeping faith with stocks
Credit Suisse Global Investment Returns Yearbook 2010 - A focus on emerging markets
Credit Suisse Global Investment Returns Yearbook 2011 – A focus on bonds
Credit Suisse Global Investment Returns Yearbook 2012 – Inflation and currencies
Credit Suisse Global Investment Returns Yearbook 2013 – Low return world, mean revision, inflation
Credit Suisse Global Investment Returns Yearbook 2014 – Emerging markets revisited
Credit Suisse Global Investment Returns Yearbook 2015 – Responsible investing
Credit Suisse Global Investment Returns Yearbook 2016 – Interest rates and investment returns
Credit Suisse Global Investment Returns Yearbook 2017 - Inflation, bills and bonds over the long run
Credit Suisse Global Investment Returns Yearbook 2018 – Risk premiums and factor investing
Credit Suisse Global Investment Returns Yearbook 2019 – Emerging markets
Credit Suisse Global Investment Returns Yearbook 2020 – ESG investing
Lessons from recent history
Five lessons from history
The Madoff Affair
The Trillion Dollar Bet BBC
Links to research and analysis
Journal of Finance
Journal of Financial Economics
Eugene Fama
AQR Library
Bill Bernstein
Jeremy Grantham
- Sample portfolio with slightly different factor exposure but similar long-term expected return