2/ "The starting point is simple yield measures. For equity assets, we use trailing E/P. For bond assets, we use standard yield-to-maturity. Both are adjusted for local risk-free rates. (For bonds, we are effectively taking the term premium as our valuation indicator.)"
3/ "We then apply asset-specific, fixed adjustments to yields. These were chosen such that the main structural biases towards certain asset classes are removed."
'Combined' rankings weight value by 50% and momentum by 50% (split equally between 1 month and 12-1 month momentum).
4/ "The two momentum strategies exhibit a positive correlation of 0.3 with each other and negative correlations of -0.1 to -0.3 with the value strategy.
"Given the correlations, it is not surprising that the combined strategy is superior to each of the underlying strategies."
5/ "The combined value-momentum GTCAA strategy does not have a large structural bias towards one specific asset class.
"We also calculate for each portfolio its average net exposure to each of the twelve asset classes (measured ex-post over the entire sample period).
6/ "By subtracting the returns of these reference portfolios from the returns of the original portfolios, we adjust the latter for possible systematic biases towards certain asset classes. The Q1-Q4 return of the combination strategy remains at 11% with a t-stat larger than 5."
7/ "UMD can be mimicked with our cross-market 12-1 month momentum strategy. This is interesting: UMD is difficult to capture, as it involves frequent trading in hundreds of stocks.
"The α of the GTCAA combined strategy is robust to CAPM, 3-factor & 4-factor model adjustments."
8/ "The combined strategy survives a realistic level of transaction costs.
"It is likely that the return can be improved in practice using more sophisticated buy/sell rules or a portfolio optimization to trade off gross expected returns against transaction costs."
9/ "In a pre-sample test from January 1974 to January 1986, returns of the various GTCAA strategies tend to be somewhat lower but still strong. (Data for emerging equities, US HY bonds, German government bonds and Japanese government bonds are not available over this period.)"
10/ "We take for each asset its annualized volatility over the past 60 months and use this to lever or de-lever the position in that asset over the next month to an (arbitrary) target volatility level of 10%.
"The results are robust to this choice of volatility targeting."
11/ "Most asset classes earn an average excess return of at least 0.3% (at most 0.2%) during months they are top (bottom) ranked, with a median of 0.8% (zero) across different asset classes.
"Valuation & momentum seem to be driven by all asset classes rather than only a few."
12/ "We cannot rule out a risk-based explanation, but it seems unlikely.
"There is no evidence of increased risk for the high return top-quartile portfolio compared to the low-return bottom-quartile in terms of volatility, skewness and other measures (statistics not reported)."
13/ "Results are not consistent with a time-varying risk explanation of our results, as alpha spreads are positive in all states of the world."
14/ "We lack a solid theoretical framework for pricing a heterogeneous set of asset classes.
"Allocation decisions across classes tend to be made by end-investors such as pension fund boards & individuals.
"It is not hard to imagine mispricing arising at the asset-class level."
15/ "The returns of certain hedge funds may be related to our strategies. However, this is mainly driven by GTCAA 12-1 month momentum, which is strongly related to UMD.
"We even find a negative relation between the combined strategy & the aggregate CSFB/Tremont hedge fund index.
16/ "Although some hedge funds may indeed be trying to exploit some of the cross-market allocation alphas documented in this paper, the overall results do not indicate that this is occurring at the large scale which would be needed to arbitrage away all these effects."
1/ The Wealthy Renter: How to Choose Housing That Will Make You Rich (Alex Avery)
"As the biggest expense of our lives, how well we manage the cost of our housing has a greater impact on our cost of living than any other factor." (p. 7)
2/ "Our housing choices determine how long it takes to get to work, how much time we spend with friends and family, where our kids will go to school, how well we do our jobs, how much money we have for other things (travel, cars, clothes, jewellery, electronics, & collectibles).
3/ "It’s the biggest factor in when/how we retire.
"We’re conditioned to want beautiful places to live. Homeownership is aggressively marketed. Pro–home ownership policy is so pervasive that it has become part of the fabric of our society. It’s enmeshed in our belief system.
1/ Polarizing impact of science literacy and numeracy on perceived climate change risks (Kahan et al.)
"Those with the most science literacy & reasoning capacity were not most concerned about climate change; they were the group w/ the most polarization."
"We instructed subjects to rate the seriousness of climate change risk."
SCT = Science Comprehension Thesis: "Members of the public do not know/think as scientists; they fail to take climate change as seriously as scientists believe they should."
3/ CCT = Cultural Cognition Thesis: "Individuals tend to form perceptions of societal risks that cohere with values characteristic of groups with which they identify. Members are motivated to fit their interpretations of scientific evidence to competing cultural philosophies."
* Trend following as an inflation hedge
* The Russell 2000 is mainly comprised of companies that are losing money
* Portfolio construction for small investors
* Optimization is often counterproductive
1/ Nonreplicable publications are cited more than replicable ones (Serra-Garcia, Gneezy)
"The difference does not change after publication of failure to replicate. Only 12% of post-replication citations of nonreplicable findings acknowledge the failure."
2/ "Three influential replication projects tried to systematically replicate the findings in top psychology, economics, & general science journals. In psychology (economics, Nature/Science), 39% (61%, 62%) of the experiments yielded significant findings in the replication study."
3/ "The relative effect sizes of findings that did (not) replicate were 75% (close to 0%) of the original ones.
"Prediction markets, in which experts in the field bet on results before the replication studies, showed that experts could predict which findings would replicate."
1/ Mapping Investable Return Sources to Macro Environments (AQR)
"Style premia have less macro exposure than do asset classes. Additionally, a diversified portfolio (for asset classes and style premia) may rely less on a specific macroeconomic outcome."
2/ "We must stress the limitations of this type of analysis. Any empirical result is specific to the sample period (here 1972-2013) and dependent on design choices.
"Long/short returns are scaled to target 10% annual volatility. We subtract no trading costs or fees."
3/ "The weak relation between equities & growth reflects the forward-looking nature of equity returns. The correlation between annual equity returns & our contemporaneous (*next* year's) growth indicator is 0.24 (0.50).
"Styles were positive in both up and down
environments."