Asset–liability modelling and pension schemes: the application of robust optimization to USS

This paper uses a novel numerical optimization technique – robust optimization – that is well suited to solving the asset–liability management (ALM) problem for pension schemes. It requires the estimation of fewer stochastic parameters, reduces estimation risk and adopts a prudent approach to asset allocation. This study is the first to apply it to a real-world pension scheme, and the first ALM model of a pension scheme to maximize the Sharpe ratio. We disaggregate pension liabilities into three components – active members, deferred members and pensioners, and transform the optimal asset allocation into the scheme's projected contribution rate. The robust optimization model is extended to include liabilities and used to derive optimal investment policies for the Universities Superannuation Scheme (USS), benchmarked against the Sharpe and Tint, Bayes–Stein and Black–Litterman models as well as the actual USS investment decisions. Over a 144-month out-of-sample period, robust optimization is superior to the four benchmarks across 20 performance criteria and has a remarkably stable asset allocation – essentially fix-mix. These conclusions are supported by six robustness checks.     

A comparative analysis of the work values of Greece's ‘Millennial’ generation

This study adds to the growing body of international evidence concerning the millennial generation (i.e. those born after 1980), by documenting the work values of millennial-aged undergraduate students (N = 649) in Greece, using an adapted version of the Lyons Work Value Survey. Our results show that millennial-aged Greeks place most importance on intrinsic and social aspects of work, and less importance on extrinsic and prestige values. We statistically compare our results to those of published studies of millennial-aged respondents from other countries and find that Greek respondents hold a unique work value profile, although they demonstrate some similarities to counterparts in other countries. Our findings do not clearly support the notion of a ‘global youth generation’, as young people in various countries hold different work value priorities. This suggests that employers seeking to recruit, engage and retain young workers internationally must tailor their offerings to the specific cultural context.     

Harmful diversification: Evidence from alternative investments

Alternative assets have become as important as equities and fixed income in the portfolios of major investors, and so their diversification properties are also important. However, adding five alternative assets (real estate, commodities, hedge funds, emerging markets and private equity) to equity and bond portfolios is shown to be harmful for US investors. We use 19 portfolio models, in conjunction with dummy variable regression, to demonstrate this harm over the 1997–2015 period. This finding is robust to different estimation periods, risk aversion levels, and the use of two regimes. Harmful diversification into alternatives is not primarily due to transactions costs or non-normality, but to estimation risk. This is larger for alternative assets, particularly during the credit crisis which accounts for the harmful diversification of real estate, private equity and emerging markets. Diversification into commodities, and to a lesser extent hedge funds, remains harmful even when the credit crisis is excluded.     

The effect of major adverse renal cardiovascular event (MARCE) incidence,procedure volume,and unit cost on the hospital savings resulting from contrast media use in inpatient angioplasty

Objective: To determine the net economic impact of switching from low-osmolar contrast media (LOCM) to iso-osmolar contrast media (IOCM; iodixanol) in patients undergoing inpatient coronary or peripheral angioplasty in the United States (US).

Methods: A budget impact model (BIM) was developed from a hospital perspective. Nationally representative procedural and contrast media prevalence rates, along with MARCE (major adverse renal cardiovascular event) incidence and episode-related cost data were derived from Premier Hospital Data (October 2014 to September 2015). A previously estimated relative risk reduction in MARCE associated with IOCM usage (9.3%) was applied. The higher cost of IOCM was included when calculating the net impact estimates at the aggregate, hospital type, and per hospital levels. One-way (±25%) and probabilistic sensitivity analyses identified the model’s most important inputs.

Results: Based on weighted analysis, 513,882?US inpatient angioplasties and 35,610 MARCE cases were estimated annually. Switching to an “IOCM only” strategy from a “LOCM only” strategy increases contrast media cost, but prevents 2,900 MARCE events. The annual budget impact was an estimated saving of $30.71 million, aggregated across all US hospitals, $6,316 per hospital, or $60 per procedure. Net savings were maintained across all univariate sensitivity analyses. While MARCE/event-free cost differential was the most important factor driving total net savings for hospitals in the Northeast and West, procedural volume was important in the Midwest and rural locations.

Conclusions: Switching to an “IOCM only” strategy from a “LOCM only” approach yields substantial net global savings to hospitals, both at the national level and within hospital sub-groups. Hospital administrators should maintain awareness of the factors that are likely to be more influential for their hospital and recognize that purchasing on the basis of lower contrast media cost may result in higher overall costs for patients undergoing inpatient angioplasty.     

Designing feasible and effective health plan payments in countries with data availability constraints

Risk equalization schemes, which transfer money to/from insurers that have above/below average risks, are a fundamental tool in regulated health insurance markets in many countries. Risk sharing (the transfer of some responsibility for costs from a plan to the regulator or the overall insurance market), are an additional method of insulating insurers who attract higher-than-average risks. This paper proposes, implements and quantifies incorporating risk sharing within a risk equalization scheme that can be applied in a data-poor context. Using Chile's private health insurance market as case study, we show that modest amount of risk sharing greatly improves fit even in simple demographic-based risk equalization. Expanding the model's formula to include morbidity-based adjustors and risk sharing redirects compensations at insurer level and reduces opportunity to engage in profitable risk selection at the group level. Our emphasis on feasibility may make alternatives proposed attractive to countries facing data-availability constraints.     

Towards a topic modeling approach to semi-automatically detect self-reported stroke symptoms (FAST symptoms) and their correlation with aphasia types

Quality & Quantity - This paper examines whether stroke survivors express that they suffered stroke symptoms on a question from the self-reported informal part of comprehensive test for...