Science topic

# Portfolio Optimization - Science topic

Explore the latest questions and answers in Portfolio Optimization, and find Portfolio Optimization experts.

Questions related to Portfolio Optimization

Dear scientific community,

I would be very interested to hear your input regarding the scaling-up of LCA studies to a portfolio level. I know there is a plethora of product LCAs and plenty of them consider several individual products or product variants in parallel. However, I have not found an awful lot of studies that extend to several hundred, let alone thousands of individual products within the scope of one study (as opposed to equally as many individual case studies).

Surely, more people have approached this apparent research gap. So for anyone that has been active in this area: I would greatly appreciate you sharing what experience you have made or you pointing me at any related publications in the field.

Many thanks and best regards

Tobias

P.S: If you are interested what my colleagues and I have done in this field, feel free to check this framework article and the case study we presented at LCM 2021 conference:

Conference Paper The Sustainability Data Science Life Cycle for automating mu...

Conference Paper Detecting environmental hotspots in extensive portfolios thr...

Hi Friends,

I would like to make an article in Quantum Finance.

You know that computer giants such as IBM and D-Wave System have opened quantum computers over the cloud to the use of researchers, currently.

I want to do Quantum portfolio optimization using Quantum computers using different financial assets (Stock, currency, gold, crypto).

I have tried with D-Wave System but I still haven't been successful:/ It is really diffucult.

Can you help me or give me an idea about this?

Thank you very much

Best Regards

I am working on portfolio optimization using lower partial moment of order 1, can someone help me how to implement LPM-1 in excel sheet using "tau" as my threshold value as 0.00% and order (n) as 1.

Thank you all in advance for your contributions to my question.

For orientation I am looking for theories as a basis of an goal-based investment approach. So far I found some appraoches using

- Portfolio Optimization with Mental Accounts (POMA)
- Asset Liability Management (ALM)
- Stochastic Dominance (SD)

Do you have any other suggestions?

Which approach is most common and accepted?

What is the best way to construct investment portfolios: based on asset prices or based on the changes in asset prices through time? E.g., if Asset A is 55 in 2010 and 60 in 2011, should I use 55 and 60 as data points or should I use 5 (i.e, 60 minus 55)? Thank you in advance.

Hi,

I want to calculate the correlation of the Bitcoin price with some other asset classes such as gold or oil. Therefore, I use the daily returns in percent of each asset and simply apply the correlation formula in excel. To be sure I got me two different datasources for oil and gold (Krugerrand and H&H gold, Texas and London Oil). Obviously the historic prices of Krugerrand and H&H are almost the same, the same applies for the two different kinds of oil.

My question/problem is: The correlation of the absolute values from Gold 1 - Gold 2 and Oil 1 - Oil 2 is close to one which makes sense. The correlation of the returns, however, is close to zero, i.e. non existent. How can that be? shouldn't the returns be at least very positively correlated as well?

Furthermore, the correlation of each time series with Bitcoin differs, even though I am thinking that Oil 1 and Oil 2 / Gold 1 / Gold 2 should have the same relation to Bitcoin.

I am trying to figure this out since days now and I am running out of time for this project. I would really appreciate if anyone has an idea what I am not seeing here.

I attached the excel, the most relevant sheets are highlighted in red.

**Question**

- Is there a benefit of having lower gap between 'in-sample' variance of portfolio daily returns and 'out-of-sample' variance of portfolio daily returns? (= better estimates the out-of-sample variance)

**Question in more detail**

- I have developed a way of optimizing a portfolio, based on Global Minimum Variance portfolio optimization.

- There are upside and downside of my portfolio optimization method.

cons : It cannot lower the `out-of-sample variance' of portfolio daily returns than the GMV portfolio. In other words, my portfolio optimization method fails to achieve better portfolio performance such as Sharpe ratio.

pros : However, my portfolio optimization method has lower gap between 'in-sample variance' and 'out-of-sample' variance than the GMV portfolio.

- For illustration, let me give you an example. During training period to come up with how much weight to put on each stock, GMV portfolio optimization calculates the stock weight with variance 100 of daily returns. However, during the investment period (test period), it gives me 125 for 'out-of-sample' variance.

- My portfolio method gives me 150 variance for 'in-sample variance' and 130 for 'out-of-sample' variance. As you can see, the actual variance is still low with the result of GMV portfolio optimization method. However, my method expects the 'out-of-sample' variance better than GMV method. GMV method is wrong by 25 percent, while my method is wrong by 13 percent.

- As such, I am curious to know if my portfolio optimization method would be useful in any case of trading in stock market nowadays.

How do you approach the selection and implementation of an information system for managing projects, programs and portfolios in an organization? What are the main criteria for choosing and implementing a management information system you can recommend?

Price optimization methods and algorithms are used to determine the best price or set of prices for business offerings by companies. In our project https://www.researchgate.net/project/Dynamic-Pricing-Algorithms-and-Models-using-Artificial-Intelligence

We are working on Dynamic Pricing Algorithms and Models using Artificial Intelligence. However we would like to hear from researchers experts about dynamic pricing models and algorithms. What are the best of breed Dynamic Pricing Algorithms and Models using Artificial Intelligence?

I am an undergraduate student and in a project I am using copula to get multivariate distribution and use this distribution to get optimal weights of my portfolio in different risk measures(VaR/ CVaR/ Draw-downs). Although, there are enough research paper available but It would be really helpful if someone please suggest me some

**Programming in R for risk measures(VaR/ CVaR/ Draw-downs)**.Most prefered:- Investment optimization as an extension of Unit Commitment for a Portfolio consisting energy generation Units.

- Unit commiment

- Investment Optimization

- Economic dispatch

- Long Term investment decision.

- Maximize gains

- Main constraint is to fulfill the demand

- Capturing all the uncertainities.

- Deterministic.

- Energy economy dispatch.

Mixed-Integer Linear Programming,

Exact algorithm

The project is to maximize the Net Present Value of Heat generating power plants in a District heating system.

- Using Mixed integer liner programming.

- I need to determine the optimal timing to invest(maximizing Net Present Value) in district heating power plants and at the same time minimizaing Carbon emissions.

- Main constraint: coverage of a given heat demand.

- Investment decision through mixed integer linear programming.

- Investment optimization as extension of unit commitment. (schedule optimization)

- Deterministic approach.

- Investment optimization as an extension of unit commitement Optimization(Schedule Optimization).

- District Heating

- Revenue streams are Heat and Electricity.

- Main Objectives:- Optimization of Investment and Carbon emission

- Main constraint: coverage of a given heat demand.

Using Mixed Integer Linear Programming

1) Main constraint: coverage of a given heat demand.

2) Investment decision through mixed integer linear programming.

3) Investment optimization as extension of unit commitment (schedule optimization)

I am working on a project in which I need to optimize the scheduling of heat generating power plants.

I want to Maximize Net Present Value and at the same time minimize Carbon emission.

Mixed integer liner programming will be used through out the modelling.

Does anyone have research insights on whether Millennials and their investment needs and habits differ from those of their parents? And if so, how they differ?

I'm looking for some advice for a suitable technique for optimising subject to constraints. My initial problem started out as a basic simplex optimisation problem finding a weight vector $w$, but now I'm trying to add more special constraints. I've heard of techniques to transform absolute function constraints into multiple linear constraints, but I don't know about sign functions. Here is my problem: given input values $x_{i}^{j}$ and $y_{i}^{j}$, I'm trying to find a weight vector $w_{i}^{j}$ to optimise:

\[ \sum^{m}_{j=1}\sum^{n}_{i=1}(w_{i}^{j}+1)y_{i}^{j} \]

subject to:

\[ \sum^{m}_{j=1}\left| \sum^{n}_{i=1}w_{i}^{j}x_{i}^{j}\right| \leq H \]

and

\[ -1 \leq w_{i}^{j} \leq 1 \] for all $i$ and $j$

and

\[ \left|\sum^{n}_{i=1}w_{i}^{j}x_{i}^{j}\right|\leq \left|\sum^{n}_{i=1}x_{i}^{j}\right| \] for all $j$ and

and

\[ sign\left(\sum^{n}_{i=1}w_{i}^{j}x_{i}^{j}\right) \left(sign\left(\sum^{n}_{i=1}w_{i}^{j}x_{i}^{j}\right) + sign\left(\sum^{n}_{i=1}x_{i}^{j}\right) \right) = 0\] for all $j$

Any advice anyone can give would be greatly appreciated, I dont really want to go down the Genetic Algorithm route straight away, if theres a numerical programming technique that is suitable I'd prefer to work with that and if this can be transformed into basic linear constraints and I can carry on using simplex optimisation that would be great!

I'm trying my hand at this problem.

I have two portfolios created with different strategies and different samples of the same investment universe. How can I determine if performance differences are statistically significant? I have several tests available to evaluate efficient frontier with ex-ante data (Spanning test; or Gibbons, Ross, and Shanken (1989) - those I prefer).

My problem is the evaluation of ex-post performance.

Can I still use Gibbons, Ross, and Shanken (1989)?

A simple test of difference in the means (t-statistics)?

Thanks to everyone who can help me, thanks for your time.

I'm interested in financial economics and also I have ability to work with Artificial neural network and fuzzy logic.

is there any article about clustering or classification with financial approach?

Have you any idea about application of clustering and classification in financial markets?

I want to optimize a portfolio based on regime switching model. Before optimization i need to detect the hidden regimes in the data. Let us suppose i want to start my analysis with GDP. What i want is to know the steps that are involved in identifying the regimes in a variable. There are many packages which do it.

The first step is to download data for GDP. Then how should i proceed to detect the regimes. I will appreciate if some one just list the steps. For now i am not interested in the model or calculation. I just need to know what are the general steps in detecting the regimes in a variable.

I want to optimize a portfolio in different horizon times.

can you help me what can I do and which software be suitable for this kind of problem? especially about portfolio re-balancing?

For example, if we have returns of two assets A and B for six days, how can we calculate an equally weighted portfolio of these assets.

**Time period Return of Asset A Return of Asset B**

Day 1 -0.710642873 -5.393463923

Day 2 -0.710642873 -5.393463923

3 -1.095970421 -8.495258637

4 1.127622947 2.971853106

5 1.417992995 -2.789204653

6 0.419478214 -2.027719244

please also share a relevant article of any good journal for reference.

Regards,

Naveed

I'm currently doing my portfolio analysis with 30 stocks. I use FCM to cluster my data and use GA as my portfolio optimizer. I have 3 clusters. Every time I run my code in GA, my plot looks like the file attached below. It's ok to have an efficient frontier like that?

Hey,

I'm currently looking for a scenario tree modeling the return of some stocks or indices. As scenario tree generation is not the focus of my work, but I need some to test my portfolio model, it would be nice, if someone could send me his work. There is no need that the data is up-to-date. Ideally the tree has 3-5 stages and 3-5 stocks/indices.

Thank you very much!

Jelto

I'm currently working on PORTFOLIO Optimization for my research paper. I used GA to solve a multi-objective problem but instead of using binary number as a gene indicator/ chromosome I used real number. Then I normalized it for the weights to be equal to 1. Is it ok to use real numbers instead of binary numbers?

Hey,

I'm looking for a way to model different kinds of transaction costs (bid-ask spreads, brokerage commissions, market impact costs, transaction taxes) in a multi-period risk minimimzation model.

In papers dealing with multi-period portfolio optimization you can often find proportional transaction cost, which doesn't seem to be very realistic. Do there exist good alternatives which lead to a "solvable" optimization problem? Are there any best practices from active traders available?

Thank you!

Jelto

Hey,

in the moment I'm dealing with time consistency for optimization problems in the sense of Shapiro(2009). My problem in the moment is, that I think that a minimal expected return constraint for the final wealth normally leads to time inconsistent problems but I cannot find any papers about this topic. Carpentier et al. (2012) adresses this question in the Markovian setting, but I cannot find anything for the general one.

This leads to my final questions: Is there some literature adressing the question, how far minimal expected return constraints have effect on the "degree" of time inconsistency of the optimization problem? Are there similar formulations (for example for intertemporal return) that are mitigating this problem?

Thanks for your help.

Jelto

I am still struggling to understand the exact differences between Multi-attribute Decision Making (MADM) and Multi-objective Decision Making (MODM). From what I know a major difference is the solution space (discrete vs continuous).

Could someone provide me with a practical example to illustrate that difference? An (project) portfolio selection example would be even more helpful.

Thanks a lot!

Hey everybody,

I'm dealing with time consistent portfolio optimization (as done in Yang,Cheng, Zhang 2015) and I want to show, that the nested CVaR defined for a wealth process (X_t, ..., X_T) by

ComCVaR_{T,T}(X_T) = - X_T

ComCVaR_{t,T}(X_t,..., X_T} = CVaR(-ComCVaR_{t+1,T}(X_t+1,..., X_T}| F_t) - X_t

leads to inconsistent policies in the sense that the future planned decisions are not going to be implemented when running the optimization problem (Min ComCVaR) at a later time stage. Are there any easy examples in the form of scenario trees which show the time inconsistency of the optimal policy in the literature or can they be constructed easily? Have not found any up to now.

Thanks for your help!

Jelto

Graph algorithms like connectivity,search etc. I have read research work on MST algorithms and work of Mantega (Hierarchical structures in Financial Markets).

Is there already some research going on, trying to generalize the work from Rockafellar and Uryasev (Optimization of conditional value-at-risk) to non-coherent risk measures induced by OCEs? The only thing I found was a paper from Natarajan, Sim and Uichanco (Fractable Robust Expected Utility and Risk Models), who are using piecewise linear utility functions. But it seems difficult to use this for larger problems, or am I wrong?

Looking forward for getting some interesting input!

Jelto

If I have one year of daily closing prices of S&P 500 stocks (the first 100, alphabetically by ticker, with a full year of data from January 9, 1998 to January 8, 1999), then how can I scale it for a portfolio holding period of 20 days?

when i run CVaR portfolio model in matlab i can just see final variable but i also need simplex final tableau.

I just have a Electre tri software.

I'm going to compare some sorting methods (around 5) in a stock portfolio problem. then I should find some proper sorting methods and softwares.

As both the mu vector and the S matrix suffer from estimation errors Robust portfolio optimization and robust portfolio estimators is a subject of interest. How can the Mahalanobis Distance be applied in portfolio optimization theory?

Does out of sample performance of minimum variance portfolios improve when the estimates of the optimizer is a high dimension covariance matrix with high-frequency data and realized volatility approach?

Using the historical var-cov matrix as an input in the optimizer leads to estimation errors. What other methods can be used in estimating the var-covar apart from shrinkage and diagonal methods?

I am considering which algorithm to use for Multi-class classification of credit risk scores.

I have a training set of ~100 default rates... and another transaction history of 1000 companies.

The idea is to classify in 5 groups A, B, C, D, E

My plan was to use Support Vector Machine for Mult-class. but I am open for new ideas.

Thanks