Portfolio Optimization - Science topic

Hello Dear Imran,

Were you able to find the code of steps for this methodology? Could you please share if you did?

Tobias Manuel Prenzel Cradle-to-gate is an evaluation of a portion of a product's life cycle from resource extraction (cradle) through factory gate (ie, before it is transported to the consumer). Cradle-to-gate evaluations are occasionally used as the foundation for business-to-business environmental product declarations (EPDs).

Interesting topic.

Try to look for it here:

Based on the substitution of personal goals for market goals - these are more theories from the field of behavioral finance. As we say: a tit in the hand is better than a crane in the sky!

Jan Jansen there is a chapter on "Fractionally Differentiated Features" in Lopez de Prado's "Advances in Financial Machine Learning". It deals with the trade-off between making time-series stationary by using returns and the memory-loss that is generated by doing so.

The calculation of the correlation is done based on the time invariant data pairs (xi,yi). For the calculation of the returns the derivative between 2-time steps is considered, right. This is in principle something different and here we see a nice example that this can lead to interesting results.

The correlation of the day by day return is zero. Conclusion: there is no intraday correlation between these two values.

But there might by a correlation between day i and day i+n. So try to correlate (xi,yi+1), (xi, yi+2),… and you will observe an interesting result 😉. Hope that helped – and good luck.

I agree, you should test if the differences are statistical significant with a chi-squared test. I guess, probably they are not, according to my experience.

Krystsina Mazhei great question. The PMO often includes portfolio management responsibilities along with a team of project managers actually carrying out projects. It's more than establishing rules and guidelines for project management, the PMO should be delivering value.

Based on over 10 years of work with Fortune 500 companies, the biggest mistake PMO's make when selecting a portfolio management system (PPM) for their PMO is not accounting for their current state maturity. The failure rate of these tools is high, and while there are a few key reasons for failure, part of the problem is the lack of user adoption due to a complex tool.

According to research by Gartner, 80% of PMO's are only level 1 or level 2 maturity (out of a 1-5 scale on the capability maturity model). Unfortunately, most PMO's select software that is more sophisticated than what the PMO can actually use.

I have written more about these problems in an article here:

https://acuityppm.com/the-current-state-of-ppm-software/

Refer this link... https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/746353/Algorithms_econ_report.pdf

The best reference you can have is McNeil, A. J., Frey, R., & Embrechts, P. (2005). Quantitative risk management: Concepts, techniques and tools (Vol. 3). Princeton: Princeton university press. In chapter 7. Programming in R that measures is not a soft problem but, you can look at my last paper in Risk Management journal, Cardona, E., Mora-Valencia, A. & Velásquez-Gaviria, D. Risk Manag (2018). https://doi.org/10.1057/s41283-018-0046-z.

It may can help you:

I'm a big fan of the IRR approach for ranking projects. But none of these cash flow methods takes into account the opportunity costs of negative externalities.

Hi Manish Khanra !

Try find and read recommendation about project portfolio planning from PMI Standard for Portfolio Management (overview is available at https://www.pmi.org/learning/library/pmi-standard-portfolio-management-8216).

Here are some arguments for/against some of the methods:

1) For the PV approach and perpetuity, you would need a discounting rate. The assumption of a constant discounting rate is often not true. In the case of power plants, you would probably discount some benefit amount in dollars (which is a random variable) with some interest rate (a random variable as well).

2) The PV and perpetuity assume that the power plant will not burnt down (during n periods or in perpetuity), which is surely impossible. Any potential risk and uncertainty are usually assumed to be zero in both of the methods.

3) Often these projects are decided by politicians, who cannot think over their legislation periods. So perpetuity method is surely not ideal for these particular cases.

4) Annuity only shows information from one single period and hinders differentiation between high-risk and low-risk investments.

5) Assuming risk and interest rate volatility are not issues, perpetuity shows the net asset value the best. Ownership can be transfered at any time with the same price. PV would assume that the power plant will be priced with 0$ after n periods.

In short: Each approach has assumptions. You may prefer one to the other, if you are sceptical about the assumptions in your case. On the other hand, if you do not take any assumption, your calculation may only contain minimal information.

Thanks a lot Mr. Temitayo Bankole !

Dear J. W. Nieuwenhuis ,

How a model using Mixed Integer Linear Programming can be made for 'Real Option Analysis' for an Investment decision?

I hope now it makes sense.

Dear sir

I have many papers about dispersed power generation included microturbine, fuel cell, photocell,.... you can see it in my page in researchgate.

IRR and NPV are best suitable methods

I think you can use global algorithm that observe all of equations

Well, first you should find an equation or a function in order to find the value of two variables

then you can use GA algorithm or PSO algorithm

please check the following links

Dear Patrick

Some studies address your researchgate question:

Thank you very very much guys, apologies for my vagueness I meant to say its maximise, its my first attempt using Latex since my PhD many years ago. Thank you for the suggestions, my only real experience of operational research was a 2nd year course during my maths undergrad where I learnt about the simplex method. I had a quick look at CPlex it does indeed have a lot more functionality than the standard Microsoft Solver Foundation I'm currently using. As you mentioned I am just using those sign functions like logical or statements, I didn't even realise you could use logical constraints like that, could be very powerful for my research and far more intuitive for my applications.

.

Dear Alberto,

Please refer to one of my latest coauthored papers published in an A* ranked journal, European Journal of Operational Research (Attached Within):

Al Janabi, Mazin A. M., Arreola Hernandez, Jose, Berger, Theo, Khuong Nguyen, Duc, “Multivariate Dependence and Portfolio Optimization Algorithms under Illiquid Market Conditions”, European Journal of Operational Research, Vol. 259, No. 3, pp. 1121-1131, 2017. [Publisher: Elsevier, Inc.]

I hope it is useful!

Best Regards

Tecnologico de Monterrey, EGADE Business School,

Santa Fe Campus, Mexico City, Mexico.

You might also find the following article written by Shaker A. Zahra useful:

Environment, corporate entrepreneurship, and financial performance: A taxonomic approach, Journal of Business Venturing

Volume 8, Issue 4, July 1993, Pages 319-340

Suppose that the GDP is observed at fixed intervals of time of length ∆. Let the total number of observation is N. Let S(i) be the observation at the end of i-th interval. Now calculate the return

Choose n much smaller than N. A standard choice is n=20 whereas N is of the order 10^3 or more. For every i>n, calculate the moving mean ̄r(i) and moving standard deviation sd(i) of {r(i-1), r(i-2), ..., r(i-n)}

using

The empirical drift μ(i) and volatility σ(i)  are given by

Thus one obtains two different time series {μ(i)} and {σ(i)} respectively. The range of these are divided into finitely many disjoint parts, called regimes. After you decide your partitioning, for every i, look up where do μ(i) and σ(i)  belong. Furthermore, for every sub-interval, you may assign a representative value. 

I think LINGO

you can add all return for Day 1nad divide by number of assets in Day 1...i.e. 2

If you want to you can take alook at my joint published paper on the Banking Sector Asset Management on my RG page, it shows how you can model and compute the results using fractal geometry for factorisation in four dimensions - 3 plotted, 1 visual which is relativistic and follows the sytem cybernetic resolution presrving the "Arrow of Time" postulate.

Earl Prof. Dr. SKM QC EPS Fellow (Indirect) MES MRES MAICTE

I am working on it and I am almost done with it just wait for few months. Other scenarios are there in literature which can guide you for asset  return. You can  use bound and bound algorithm and can try yourself.

Charlene Mae De vera Galang

If you use real number you will get more search space as compared to the binary one.

please go through the my publication for more detail or use the Goldberg's book.

Linear (on trading volume) market impact model IS realistic with the given time of execution. Please take a look at my paper "Market Impact Paradoxes".

I implemented multi-period portfolio liquidation scheme for Deutsche Bank and later provided short description in another paper " Statistical  Arbitrage: medium Frequency Portfolio Trading". So, it is a real model that works.  

Dear Jelto Borgmann

Greetings,

Please find the three attached files about the literature review about your question related for your topic may be useful for you and good luck

Best Regards,

For simplicity, I will take only three attributes for portfolio selection. Let there be five IT companies c1,...,c5 and three attributes, namely profit, debt equity ratio and eps. We collect past data of three attributes for these five companies. You can also see this data as 5 points in three dimensional space. Out of these five points, we have to select the best one using some criterion. Thus, it became a problem of selection from a finite set. This is MADM.

Now consider the following problem. I have taken this problem just randomly to explain. It may not have any solution.

Max f1=x1+x2-x3;

Min f2=x1-2*x2+3*x3;

subject to

g1(x1,x2)=3*x1+4*x2-x3<=5;

-1<=x1<=1, 2<=x2<=3; -2<=x3<=4;

Between the specified range of x1,x2 and x3 and those points which satisfy g1 also, there are infinitely many points from which I want a point which maximizes f1 and minimizes f2 simultaneously. These points are also from three dimensional space but the number of points are infinite. We are not given limited alternatives. In this case we have infinite alternatives. This is MODM.

I hope now you might have understood the difference.

For your kind information, our paper on portfolio selection is in press. I will upload it when it is published.

Unfortunately, I was not so active in Q&A when you asked the question. So, too late answer.

Hey Jelto,

I remember seeing in a paper (unfortunately I do not remember exactly the author and title) an example which included a sample graph with distances between different points and the time required to travel the distances. Then some of the timings are allowed to change depending on the presence/lack of traffic jams (so depending on the point of time where you get information about traffic jam your original estimates of the best way may change).

I think a similar approach may be taken with a tree (instead of a graph) and estimated losses (instead of time) and special economic events (instead of traffic jams).

Best regards,

 Stan

If you want grand unified theory of portfolio optimisation using graphs you can read a number of my papers especially the ones on Asset Management on my RG webpage.

I hope that my working paper "Applications of Risk-Sensitive Value Measure Method to Portfolio Evaluation Problems" shall help your research.

The non-linearity of the equations coupled with the feedback process makes the resulting values tend to infinity; that is why we need a few attractors parameters to maintain the range within reasonable amounts.

What we are saying that we view thinking about the climate: all good hurricane hovers around a vortex, otherwise, the storm would disrupt and was extinguished quickly.

The simplest attractor of markets is its mean, that is, the values may disperse but tend to return to around the average. Hence, assuming that markets respond to a chaotic type model then we talk about the creation of fractals markets. This is the sensitivity to initial conditions (price ex post) is one of the attributes of chaos, the pattern is identical to that produced by a random IFS certain pairs of prices excluded

There are at least two ways to reach the goal. 1. Create portfolio optimization algorithm from stratch (in Matlab or any other language), so that you have access to all interior variables, including the final simplex tableau. 2. Use a third party portfolio optimization algorithm implemented (say) in Matlab. In the latter case answering your question requires having the third party code in front of ones eyes. Cuold you please point to the place where the code resides?

Hi Seyed,

There are several software available. Find here some pointers. Best is to contact the authors if you are interested in their method.

- Rough Sets: http://www.lcb.uu.se/tools/rosetta/#functionality

- FlowSort  sorting method (based on Promethee methodology): www.smart-picker.com  --> there is a trial version and also research licences

- Tomaso method (http://public.telecom-bretagne.eu/~pmeyer//software/tomaso/tomaso_tutorial.html)

- SMAA (http://www.inescc.pt/~ewgmcda/SW_CSMAA.pdf)

- FINCLASS (Multicriteria Decision Support for Financial Classification Problems) (in Newsletter, Series 3, No. 3, Spring 2001): http://www.cs.put.poznan.pl/ewgmcda/pdf/SWFinclass.pdf

Other references:

Best wishes,

Philippe

Dear Anirban,

My approach to the problem you describe has been to integrate a Kalman Filter for estimating the joint returns distribution (under a linear Gauss-Markov dynamics assumption) into a multi-objective optimizer, which is an anticipatory metaheuristic.

One of the most important aspects when integrating predictive knowledge into an optimizer is to define a proper and better suited optimization model.

Thus, in my doctoral research, I've devised the so-called Anticipatory Stochastic Multi-Objective Optimization (AS-MOO) model, which is a multiperiod recurrence equation explicitly incorporating predictive knowledge about the future moments of factored returns distributions, for each Pareto-efficient candidate portfolio.

Perhaps you will find the following attached publications useful.

Best regards,

Carlos.

Thanks Markus

Implied Volatility as it is only the market's prediction of it from a Black- Scholes model, with quoted option prices , need not necessarily follow the statistical properties of a variance- co variance matrix .'Volatility smile' is an example for it

There are two things that I would advise is:

a)As Sylantyev mentioned the time period is important due to crisis post-crisis period. you should test the data to see if the features correlate to the time periods. Also, see if shorter time periods have an effect on the data (spot cyclical or specific time periods which might bias the data)

b) At the stage that you are mentioning right now it is better to test several algorithms and not just one. Unless you have previous literature that already has done the work for you, I would suggest that you try several methods.