TLDR: It’s the art and science of making decisions about investments to maximize return for a given risk appetite. And it Rocks!!

The long version From Investopedia:

Portfolio management is the art and science of making decisions about investment mix and policy, matching investments to objectives, asset allocation for individuals and institutions, and balancing risk against performance. Portfolio management is all about determining strengths, weaknesses, opportunities and threats in the choice of debt vs. equity, domestic vs. international, growth vs. safety, and many other trade-offs encountered in the attempt to maximize return at a given appetite for risk.

Let’s break that down into its parts:

1. Making Decisions
2. To Maximize Return
3. For a given Risk Appetite
4. The Science of Portfolio
5. The Art of Portfolio

1.   Making Decisions

Portfolio analysis is all about making investment decisions. The idea is to make the RIGHT decisions so that you make money (i.e. maximize return). The key thing here is to understand that to make the right decisions you need the right data to support your decision making. This is the hardest part of portfolio analysis. Data discovery, collection, and standardization is the first part of any portfolio.

To do effective portfolio analysis we must understand that there is a limit to the number of decisions that a company (or human) can efficiently manage. As the company gets larger and more sophisticated, the number of decisions may increase but only up to a point. What a company should be doing is increasing the size or ‘Materiality’ of the decisions units. In the case of an O&G company, this may mean moving from making decisions at a well level to an asset level, or from a field to a block level. These need to be thought through to make sure that the opportunities line up with the decisions that the company wants to make. A super major should not be using their portfolio to make decisions on which well to invest in.

2.   To Maximize Return

Once we have a collection of opportunities then we need to build and evaluate the return of each of the opportunities. A ‘return’ could be any metric. It could be the NPV of the overall portfolio. It could be the production growth over the next five years expressed as a percentage. It could be the Reserve Replacement Ratio. If we try to maximize each of these we may get completely different outcomes resulting in a different mix of investment decisions. Depending on the company and their strategy the desired ‘return’ could be very different. This is where the art starts, but I’ll come back to that later.

So we need to model as many ‘return’ metrics as possible. These will be dictated by the company strategy and we need to allow for all of the inputs required to calculate those metrics. The added issue here is that the metrics may be dependent on when the decisions to invest are made, or to how much of an investment decision we make. In that case, we need to make sure that the return metric is responsive to the decisions being made.

The best way to make sure we can find a solution that maximizes the return is to include all of the possible opportunities in our portfolio. The list of opportunities going into the portfolio should not be filtered down to only the ones that the assets believe make sense. Doing this introduces survivorship bias into the portfolio process and does not allow the most efficient portfolios to be found.

The best portfolio analysis will contain as many possible opportunities as can be evaluated, even some which may not make sense when viewed stand-alone. Moreover, in the case of an O&G portfolio, it should contain multiple opportunities representing the same decision. For example, instead of putting a single opportunity representing the development of an asset using 3 rigs; include the 3 rig case but also a 2 rig case, a 4 rig case and perhaps a 5 rig case. This provides options in your portfolio and just as in a financial portfolio, having options in your portfolios allows you to be responsive to changing business and market conditions.

Does your head hurt yet? If so, stop here. You understand the basics of portfolio management. I generally find this is where most people get to and then stop. However, if you find this interesting, push on. It will be worth it, I promise!

3.   For a given Risk Appetite

Adding risk to your portfolio is where things get interesting. Modern Portfolio Theory (MPT), which is focused on a portfolio of investments on the financial markets, uses the variance of the asset prices as a proxy for risk. This works because the assets are highly liquid and so the price of the asset can be easily evaluated at any point in time. This is not the case for O&G assets and as such we have to take a different approach.

To use MPT concepts in the case of an O&G portfolio we need to generalize MPT to see that we can evaluate the risk on any return metric by simply calculating the variance in that metric. To do that we need to understand all of the possible outcomes for that metric at the portfolio level. Unlike a financial portfolio, there is no analytical way to do this and so we need to turn to numerical solutions.

One way to do this is through Monte Carlo sampling of the portfolio. To achieve this, we need to have multiple outcomes for each possible opportunity which can be sampled. We then take a realization of the portfolio that has been maximized for a given return metric and run a Monte Carlo analysis over it to sample all of the possible asset outcomes to establish the possible portfolio outcomes. We then use these portfolio outcomes as our ‘sample population’ to establish the sample statistics around the portfolio. One of those statistics would be the variance, our proxy for risk!

One interesting thing to note here is that if we consider variance a proxy for risk, then for each return metric we have an associated risk metric that may be very different from every other risk metric! So the ‘riskiness’ of a portfolio is very much determined by the strategic goals of the company and which ‘return’ metric that they are most concerned about maximizing.

I have only mentioned risk up to now and have stayed away from ‘Risk Appetite’ as a concept. This is because it is a fuzzy topic which is difficult to quantify. We will come back to this in the ‘Art’ section of the manifesto.

So we now have all of the building blocks we need to complete our portfolio analysis. We have the Decision Units (the opportunities), we have our Options (Mutually Exclusive Opportunities) and we have our Outcomes (Different possible realizations of the same opportunity). But what we can see here is that to do a full portfolio analysis we need an amazing amount of data. For a single decision we could have; 3 different Options (i.e. development plans) with 3 outcomes each (P10, P50, P90) and that means we need 3x3=9 sets of data to represent a single investment decision. With a portfolio of 100 investment decisions that would mean we need 900 sets of data. And if we expand this to more options and more outcomes this just grows.

The key here is to make sure the trade-offs are well understood and that the amount of data is not so large as to make the analysis impossible, or too small as to make the analysis unsolvable.

This is the problem with Excel. Many companies will try to manage their portfolio or corporate planning processes through Excel but they come up to a fundamental limit when the data set grows too large and the need for more advanced algorithms mean that they have to delve into the dark world of macros!

4.   The Science of Portfolio

The science of portfolio goes back to a seminal essay by Harry Markowitz in 1952. Since then there have been many advances in computer science such that there are many algorithms out there that can solve MPT problems with relative ease. These algorithms fit into a category of data analytics called ‘Prescriptive Analytics’. The algorithms in this category are well known but implementing them in a piece of software that adds value and that is is easy to use and responsive to many different company situations, is very difficult!

In brief, the four types of data analytics are as follows:

• Descriptive Analytics - What happened?
• Diagnostics Analytics - Why did it happen?
• Predictive Analytics - What is likely to happen?
• Prescriptive Analytics - What decisions should we make?

Descriptive and diagnostic analytics deal in the past, predictive and prescriptive deal in the future. Each one builds on the previous and so cannot be done in isolation. This ties back into the comment earlier where we need the data and models built so that we can run the portfolio through our prescriptive analytical process and come up with a series of investment decisions as our results.

Once we have run the portfolio through our prescriptive analytical process we have a series of possible portfolio scenarios as outcomes. But which one is the right one? One might make us incredibly rich if everything goes right, but what if the market shifts and price changes? If there is a chance we will go bankrupt, would we pick that portfolio? This is where the ‘Art of Portfolio’ comes into play!

5.   The Art of Portfolio

‘Art is the expression or application of human creative skill and imagination.’

Every portfolio is different. Adding or changing one opportunity in a mix of 100 opportunities could give you drastically different results, even if you follow exactly the same process step-by-step. This means that every analysis is dependent on the structure and personality of the portfolio. Good portfolio analysis takes imagination and creativity, hence why it is an art!

The possible portfolio scenarios can be visualized in any number of ways. In my experience, I have learned a new way of analyzing results through every new portfolio analysis with every client I have worked with. The visualization of the results should be done so that they are clear and easy to follow and should tell a story. Think of it like a good presentation; it should have a beginning, a middle and an end. The starting point might be the base or current view of the portfolio, the middle might be everything that is possible from the portfolio and in the end, there should be a clear set of decisions that can be shared with the organization coming out of the portfolio.

And just like in art where beauty is in the eye of the beholder, the result of a portfolio analysis will be different based on the company, the decision makers and the shareholders. The final result from a portfolio analysis should be to select an ‘Efficient’ portfolio. An efficient portfolio maximizes return for a given level of risk. There is not one efficient portfolio but rather a set of them. In the figure below each dot is a scenario of the portfolio. The efficient portfolios fall along the line indicated.

Which portfolio will ultimately be selected depends on how much risk the organization is willing to take. This is the ‘risk appetite’ of the organization and as I mentioned earlier is a fuzzy concept that is dependent on many human factors that are hard to quantify. As such, the selection of the final portfolio can never be done solely by computers and will always require a human element!

So how does this relate to Software?

Good software makes complex ideas and processes accessible to users.  Portfolio software needs to embed all of the concepts listed above and to allow the user to work through analysis in a repeatable way.  It should also allow new users to get up to speed quickly and reduce any potential for errors in the analysis.  This is what we strive for at Aucerna.

Aucerna Portfolio is the embodiment of over 20 years of experience and thousands of person-hours in Portfolio Analysis.  We are proud of what we have built and look forward to pushing the boundaries of what Portfolio Analysis in the Oil and Gas Industry will become!

I will leave you with one last thought……

What is Portfolio Not!

Portfolio is not, and will never be, a tool that you click a button and get an answer. Portfolio will always require time and effort do. There will never be an ‘easy’ button for portfolio!!

Some references for those of you who want to do additional reading:

• https://www.investopedia.com/terms/m/modernportfoliotheory.asp 2
• https://en.wikipedia.org/wiki/Efficient_frontier 1
• https://www.pwc.com/us/en/services/consulting/analytics/big-decision-survey.html 2
• https://datafloq.com/read/the-four-types-of-data-analytics/3903 1
• https://www.scnsoft.com/blog/4-types-of-data-analytics 1