Simultaneous pursuit of out-of-sample performance and sparsity in index tracking portfolios

Index tracking is a passive investment strategy in which a fund (e.g., an ETF: exchange traded fund) manager purchases a set of assets to mimic a market index. The tracking error, i.e., the difference between the performances of the index and the portfolio, may be minimized by buying all the assets contained in the index. However, this strategy results in a considerable transaction cost and, accordingly, decreases the return of the constructed portfolio. On the other hand, a portfolio with a small cardinality may result in poor out-of-sample performance. Of interest is, thus, constructing a portfolio with good out-of-sample performance, while keeping the number of assets invested in small (i.e., sparse). In this paper, we develop a tracking portfolio model that addresses the above conflicting requirements by using a combination of L0- and L2-norms. The L2-norm regularizes the overdetermined system to impose smoothness (and hence has better out-of-sample performance), and it shrinks the solution to an equally-weighted dense portfolio. On the other hand, the L0-norm imposes a cardinality constraint that achieves sparsity (and hence a lower transaction cost). We propose a heuristic method for estimating portfolio weights, which combines a greedy search with an analytical formula embedded in it. We demonstrate that the resulting sparse portfolio has good tracking and generalization performance on historic data of weekly and monthly returns on the Nikkei 225 index and its constituent companies.     

Structure and molecular conformation of tussah silk fibroin films: Effect of heat treatment

Structural changes of tussah (Antheraea pernyi) silk fibroin films induced by heat treatment were studied as a function of the treatment temperature in the range 200–250°C. The DSC curve of tussah films with α-helix molecular conformation displayed characteristic endo and exo peaks at 216 and 226°C, respectively. These peaks first weakened and then completely disappeared after heating at 230°C. Accordingly, the TMA thermal shrinkage at 206°C disappeared when the films were heated at 230°C. The onset of weight loss was monitored at 210°C by means of TG measurements. X-ray diffraction profiles gradually changed from α-helix to β-sheet crystalline structure as the treatment temperature increased from 200 to 250°C. On raising the heating temperature above 200°C, the intensity of IR and Raman bands characteristic of β-sheet conformation increased in the whole ranges of amide and skeletal modes. The sample treated at 200°C showed a spectral pattern intermediate between α-helix and β-sheet molecular conformation. The IR marker band for random coil structure, still detectable at 200°C, disappeared at higher treatment temperatures. Spectral changes attributable to the onset of thermal degradation appeared at 230°C. © 1997 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 35: 841–847, 1997     

Study on the Morphological Change of Growing Ag Crystallites by in-situ SEM Observation

The morphology of the initial growth stage of Ag thin films on Mo(110) substrate was investigated in-situ by the scanning electron microscope (SEM). A great majority of the Ag crystallites showed a truncated Wulff's equilibrium shape of the face-centered-cubic (fcc) structure with epitaxial orientation [1 0]Ag//[11 ]Mo, (111)Ag//(110)Mo with the substrate temperature of 400 °C. In-situ SEM observation during the growth revealed two kinds of particular morphological change of Ag crystallites, namely (1) rotation of an Ag crystallite and (2) reformation of the crystal faces of the Ag crystallite. However, a small number of Ag crystal particles showed Wulff-polyhedron truncated by the (100) plane parallel to the substrate surface (oriented to the 〈100〉 axis normal to the substrate surface).     

Compatibility and mesophase separation in blends of α-helical polyglutamates

Blends composed of the α-helical polymers, poly-L-glutamates [(? NHC^αHRC′O? )_n, R = ? CH₂CH₂COO? (CH₂)_m(C₆H₅] (Lm) and the corresponding D enantiomers (Dm), have been studied by x-ray diffraction and viscoelastic measurements. Binary blends of L2, D2, L3, and D3 are compatible and form isomorphous mixed crystals at all compositions, whereas other pairs, with the exception of L1/D1, are incompatible. The demixing process is described for a ternary system consisting of L1, D3, and a diluent chloroform at 40°C. The phase diagram comprises four regions, I, IA, A, and AA, with increasing polymer concentration; I: isotropic, A: anisotropic, IA: I–A biphasic, and AA: A–A biphasic. The IA biphasic gap is greater in the ternary system than in the binary ones. The high-molecular-weight component (D3) is partitioned into the A phase in the IA region. The AA separation originates from incompatibility of the polymers. The phase behavior is discussed on the basis of the Abe-Flory theory by incorporating the polymer-polymer interaction parameter.     

Photoelastic Analysis with a Single Tricolor Image

A new approach to simultaneous measurement of isochromatics and isoclinics from a single image using a tricolor light, called as the tricolor photoelastic technique, is proposed in order to enable and facilitate the analysis of time varying phenomena. Not only an apparatus of experiment but the whole analysis system and procedure are developed. Using a color digital camera and a tricolor light source under linear polarization, fringe order and the principal direction of birefringence are obtained from a piece of color image data by a single shot using the method proposed. It is emphasized that this method can be applicable to time varying phenomena in which good repeatability and reproducibility of experiments are not expected, since multiple exposures are not necessary for sufficient data acquisition in the completion of stress analysis.     

An experimental trial for the determination of the critical coagulation concentration using the sedimentation method

A method for determining the critical coagulation concentration (C _c) from the change in the transmittance of the sol with stand time after adding a coagulating agent is discussed. Potassium nitrate was used as the coagulating agent because the specific adsorption of electrolyte ions on the particle and the hydrolysis of electrolyte ions are negligible. Apparent critical coagulation concentrations,C _c ^a, of iron (III) hydroxide and silver iodide sols were obtained from the transmittance vs. potassium nitrate concentration curves for various stand times. The values ofC _c ^a decreased with increasing stand time. TheC _c ^a value obtained for the shortest stand time was closer toC _c obtained from the initial turbidity change of the sol by applying Rayleigh's law. The Hamaker constant for the particle in water was calculated from theC _c ^a value obtained at the shortest time and the experimentally determined outer Helmholtz plane potential. The calculated Hamaker constants were comparable to the theoretical values for iron (III) hydroxide and silver iodide.     

Cl NQR study of lattice dynamic and magnetic property of a crystalline coordination polymer {CuCA(phz)(H2O)2}n

Copper(II) compounds {CuCA(phz)(H₂O)₂}_n (H₂CA = chloranilic acid, phz = phenazine) having a layer structure of -CuCA(H₂O)₂- polymer chains and phenazine were studied by ³⁵Cl nuclear quadrupole resonance (NQR). The single NQR line observed at 35.635 MHz at 261.5 K increased to 35.918 MHz at 4.2 K. The degree of reduction of electric field gradient due to lattice vibrations was similar to that of chloranilic acid crystal. Temperature dependence of spin-lattice relaxation time, T₁, of the ³⁵Cl NQR signal below 20 K, between 20 and 210 K, and above 210 K, was explained by (1) a decrease of effective electron-spin density caused by antiferromagnetic interaction, (2) a magnetic interaction between Cl nuclear-spin and electron-spins on paramagnetic Cu(II) ions, and (3) an increasing contribution from reorientation of ligand molecules, respectively. The electron spin-exchange parameter ∣J∣ between the neighboring Cu(II) electrons was estimated to be 0.33 cm⁻¹ from the T₁ value of the range 20−210 K. Comparing this value with that of J = −1.84 cm⁻¹ estimated from the magnetic susceptibility, it is suggested that the magnetic dipolar coupling with the electron spins on Cu(II) ions must be the principal mechanism for the ³⁵Cl NQR spin-lattice relaxation of {CuCA(phz)(H₂O)₂}_n but a delocalization of electron spin over the chloranilate ligand has to be taken into account.     

Method of computer algebraic calculation of the matrix elements in the second quantization language

An automated method by the algebraic programming language REDUCE3 for specifying the matrix elements expressed in second quantization language is presented and then applied to the case of the matrix elements in the TDHF theory. This program works in a very straightforward way by commuting the electron creation and annihilation operators (a^? and a) until these operators have completely vanished from the expression of the matrix element under the appropriate elimination conditions. An improved method using singlet generators of unitary transformations in the place of the electron creation and annihilation operators is also presented. This improvement reduces the time and memory required for the calculation. These methods will make programming in the field of quantum chemistry much easier. © 1995 John Wiley & Sons, Inc.     

Dielectric properties of emulsions

Summary Dielectric constants of O/W emulsions were measured over a wide range of concentration and at frequencies ranging from 20 cps. to 3 mc. No dielectric dispersion due to the interfacial polarization was observed in the experimental range of frequency, while the electrode polarization was observed below 100 kc. Experimental results were compared with theoretical values for spherical dispersions. It was concluded that the dielectric constants of O/W emulsions were expressed best byBruggeman's equation over the whole range of concentration.

---

Zusammenfassung Die DK von ?l/Wasser-Emulsionen wurden über einen weiten Konzentrationsbereich und für Frequenzen zwischen 20 Hz und 3 MHz gemessen. Es war keine dielektrische Dispersion verursacht durch Oberfl?chenpolarisation innerhalb des experimentellen Frequenzbereiches zu beobachten. Dagegen wurde unterhalb 100 kHz Elektrodenpolarisation bemerkbar. Die experimentellen Ergebnisse wurden mit den theoretischen Darstellungen der Misch-DK für Dispersionen mit kugligen Teilchen verglichen. Der Verlauf der DK der ?l/Wasser-Emulsion wird am besten und über den ganzen Konzentrationsbereich durch die Gleichung vonBruggeman dargestellt.

     

Influence of surface properties of particles on their adhesion and removal

The effect of surface properties of particles on their adhesion and removal was investigated using an immersed system consisting of nylon particles and a quartz plate. The nylon particles were dyed with a reactive dye in order to change their properties and were used for the adhesion and removal experiments in comparison with undyed particles. The electrokinetic potentials of the particles were measured by micro-electrophoresis and the Hamaker constants were independently evaluated using experimental values of dispersive component of surface free energy determined by the Wilhelmy technique. The experimental results were used for the discussion of particle adhesion and removal on the basis of the heterocoagulation theory. The differences in adhesion and removal efficiencies between dyed and undyed particles were explained in terms of the electrostatic and dispersive van der Waals interaction by considering the differences in thier properties, the electrokinetic potential and the Hamaker constant, due to dyeing.