Author: David Leevan

October 8, 2019

Fundamental vs Simulation Models for Energy Portfolios

There seems to be some confusion regarding the differences between, and best uses for, fundamental and simulation models.  In part, this is the fault of the vendors, who try to sell their solution as good for everything. There is no individual model that should be used for every analytical task. 

Recently I heard a chief officer of an energy company ask why his company needs simulation-based analytics, when they already have a fundamental model?  This article is designed to answer that question.

Fundamental vs Monte Carlo Simulation models

Fundamental power system analytics are typically used to build up an analytical representation of a market (generation, transmission, load). Making assumptions about grid topology, available generation capacity, maintenance outages, weather patterns, and aggregate load, these fundamental models will compute least-cost dispatch and optimal market pricing at an hourly granularity for many years into the future.  This is no small feat and fundamental models are quite good at this. A fundamental model is best used to develop a long-term forecast of hourly prices, particularly for use to inform resource planning efforts. Of course, this “single-path“ market forecast is always wrong, but if the model assumptions and inputs are good, these models can produce useful results. 

Knowing that the single-path forecast is going to be wrong in unforeseen ways, companies will turn to simulation models to understand a portfolio’s risk, or sensitivity to uncertainty, as prices deviate from the expected forecast. The simulation model will run hundreds or thousands of market simulations and probabilistically weight the results.  This yields a distribution of outcomes where the expected value is reported as well as the size, shape and magnitude of uncertainty within the distribution. Distributions can be inspected at any time granularity (hourly, monthly, annually) on metrics like portfolio gross margin, total costs, generation, fuel burn, and more. The left side of the distribution is interpreted as the “downside risk”.  Companies use the magnitude and timing of these downside risk events to hedge their portfolio and increase certainty around future cash flows. 

Fundamental models are not suited to generate robust distributions of value and therefore are inappropriate for assessing the effectiveness of existing hedging strategies or developing new ones.

Market Price Simulation

Price simulation is the key component to any simulation framework since market prices are a primary driver of risk. Different markets have very different pricing dynamics. Hourly power price shapes vary by day-of-week, month-of-year, and pricing location. Correlations between pricing locations and across commodities (e.g., between power and gas) vary seasonally. Volatility, the likelihood of price spikes, and mean reversion rates, all vary seasonally as well.  These price dynamics can strongly influence cycling behavior for thermal assets and impact both value and risk for intermittent renewable resources.

Many portfolio managers ask about fundamental drivers such as weather, grid topology, available generators, demand, etc. In most cases, these fundamental drivers are well represented in each market’s unique price dynamics, and therefore can be captured in the price simulations.

Interestingly, a commonly used integration between fundamental and simulation models involves using the forward-looking market price forecast generated by the fundamental model as input to the simulation model.  The simulation model will then produce thousands of hourly price paths around it, yielding the distributions needed for risk and hedging analysis while remaining consistent with the fundamental forecast itself.

Physical Asset Optimization & Valuation

Many years ago, energy analysts routinely used simple “closed-form spread option” models to value power plants.  Spread option models use a price for fuel and a price for electricity, as well as the volatility of each and the correlation between the two.  In a given month, if the expected spark spread was sufficient, the power plant was assumed to run.  If not, it was assumed to shut down.  Of course, this is an over-simplification of how power plants operate.  There are dozens of physical operating characteristics and costs that “constrain” the optimal operation of a power plant.  Modern dispatch models use a very detailed set of characteristics to accurately represent an asset so that modeled output is as close as possible to true asset operations.  Most simulation and fundamental models both use these modern techniques.  The difference is the simulation model dispatches assets to hundreds or thousands of hourly price paths, while still maintaining all operating constraints.  This yields hourly/daily/monthly/annual distributions of values for generation, costs, fuel burn, gross margin allowing for more effective hedge decisions. 

Other assets, such as renewables and storage, require separate analytical methodologies but still need to be integrated into a comprehensive portfolio analysis to properly track their effects on value and risk.  Renewable assets are non-dispatchable, but their hourly production can still be simulated and combined with market price simulations to develop distributions of generation and analyze the effects of volumetric uncertainty (swing risk) on asset value.  Such an analysis provides insight into production variability, price uncertainty, and even basis risk between a generation node and corresponding settlement hub.

Unlike renewables, batteries are dispatchable and their charge/discharge operations can be optimized to market prices or retail rates. Since batteries generate no electricity themselves, their value is entirely derived by operating them intelligently relative to various sources and sinks of energy at the battery site along with their corresponding values and/or costs (e.g., customer load, co-sited renewable generation, retail rates including TOU schedules and demand charges, and wholesale prices or export compensation rates).  Other asset types can and should be included in a comprehensive portfolio analysis. 

Contract valuation, Risk Management and Hedge Optimization

At a high level, we have discussed some of the differences between sophisticated fundamental and simulation models.  Using a simulation model, we generate thousands of hourly price paths that capture market dynamics and represent a very large set of possible future outcomes. These outcomes are probability-weighted and allow reporting not only on expected (mean) prices and portfolio value, but also left-tail and right-tail risk levels.  We can use these price simulations to optimize and value our physical assets, yielding many other useful statistics to support asset management and budgeting activities.  Finally, we can also use the same price simulations to value our physical and financial contracts.  These contracts can come in the form of standard swaps & options, index deals, heat-rate-call-options, revenue-put-options, PPAs, or other exotic derivatives and structured transactions.  Each of these contracts have unique payoff functions. 

Most importantly, we can aggregate all the above analysis based on both forward and spot price simulations to forecast distributions of portfolio cash flow through time. By bringing together all of these values, we can calculate summary statistics for the total portfolio – including both physical assets and financial contracts.  This affords the analyst / portfolio manager insight into the timing of all cash flows and how they are being generated. 

By looking at the distributions of value across months, peak periods, asset types, sub-portfolios, and many other dimensions, one can understand how, when, and why risks enter the portfolio.  This leads to the final piece of the puzzle: hedging the portfolio. 

Deciding on and implementing a hedging strategy depends on many input factors.  But armed with the analysis described above, we have all the analytical components we need to measure the risk-reduction-value (RRV) of a prospective hedging strategy.  The process above has given us distributions of value through time for every item in the energy portfolio and has accounted for the interactions between all the individual components to provide a “rolled-up” total portfolio view.  We can now assess various hedge strategies or evaluate individual contracts, assets, etc. to see the net effect they produce on the entire energy portfolio. 

No other model methodology will allow this type of analysis.  Fundamental models are very useful in their own right, but their deterministic outlook has limitations for portfolio analytics. Traders, portfolio analysts and energy risk managers should be utilizing a robust simulation-based methodology for deal valuation, cash-flow analysis, asset and hedge optimization.   

Why cQuant.io?

cQuant.io is a team of experienced PhD quants and energy analysts. cQuant develops sophisticated analytic models and delivers these solutions in a cloud-based application.  Our team has extensive experience delivering portfolio & risk analytic solutions to energy companies, including IPPs, utilities, trading floors, community choice aggregations (CCAs) and other load serving entities (LSEs). 

A cQuant solution can typically be set up for less than the cost of adding a senior quantitative analyst.

David Leevan is the CEO of cQuant.io, a SaaS platform for energy analytics.    

June 6, 2019

Five Key Mistakes Companies Make When Buying Software

After 20+ years in software sales, it is safe to say that I have seen almost everything. At a recent energy analytics event, we presented the “Top 5 Mistakes” that companies make when buying enterprise software.  Amazingly, most companies make the same mistakes, over and over. Why is that?

Vendors engage in sales processes dozens of times per year, but energy companies may only buy enterprise software once every few years. Vendors have developed many ways to fool the customer.

Frequently, I am dealing with buyers that are far less experienced that I am in the software procurement process.  While I engage in dozens of buying processes per year, the buyer may only deal with a few major software purchases per career. Amazingly, buyers seem to make the same mistakes over and over. 

While I have often tried to help the buyer, their process is usually flawed from the beginning and very difficult to correct. Here you will find the top five recurring mistakes that I encounter, along with suggestions on best practices. I hope that my time spent in the trenches can help you avoid some of these common mistakes.

Software Buying Mistakes, The Context:

  • Vendors have the upper hand
  • Vendors work with dozens of buyers each year. Buyers may only have
    1-2 experiences in an entire career
  • Vendors have gamed the system
  • Vendors want high revenue, captured customers
  • Buyers want low cost, high-value, flexibility, freedom of movement

Software Buying Mistake #1

Buying software without extensive solution testing

  • Vendors often use bespoke demo version
  • Demo version often use ‘vapor-ware’ to get the sale
  • Buyers don’t find out it was vapor-ware until after the contracts are done

Your Solution: Always engage in a Proof of Concept or Pilot in the exact configuration that you will be buying.

Software Buying Mistake #2

Evaluating too few companies

  • Buyers often “lock into” only one type of solution. 
  • In every part of the energy industry there are dozens of potential solutions.

Your Solution: Spend extra time investigating a wider range of potential options.  If nothing else, you will learn more about the solution space and feel better about your original choice.

Software Buying Mistake #3

Relying on a Solution Expert, or Solution Consultant

  • Companies often bring in solution consultants to help with their selection.
  • The consultants often have a lot of experience with the major market player.
  • The consultant may be looking for a role after you buy, as a system expert.
  • They will often insure that the most expensive solution wins.

Your Solution: Understand that consultants want longer engagements.  They do not want easy solutions.  Make your own decisions!

Software Buying Mistake #4

Requesting only 3 references

  • Any vendor can find 3 references.
  • What percentage of their customers could be a reference?
  • Ask references about initial cost estimate vs actual, initial time estimate vs actual, etc.

Your Solution: Ask for ALL customers contacts and call them at random. Or, do your own homework and find companies using the solution that the vendor will not provide as a reference. 

Software Buying Mistake #5

Using a vendor for too much.  ‘Best in Class’ vs ‘Single Source’

  • Vendors love building their product at customer’s expense.
  • 20% of the project takes 80% of the time and money.

Your Solution: Use vendors for their core offering only.  But if you do decide to ask the vendor to enhance their solution, the vendor should do this for free!

David Leevan is the Managing Director of cQuant.io, a SaaS platform for energy analytics.    

Are you ready to improve your energy analytics? Request your free demo today.

SCHEDULE YOUR DEMO TODAY

October 16, 2018

Renewable PPA Primer: Financial Risk Assessment for Corporate Energy Buyers

Renewable energy purchases by large corporations are accelerating.  In 2017, large corporates bought over 5 gigawatts of renewable energy and we are on-pace to shatter this number in 2018 with over 7 gigawatts already purchased this year (Bloomberg article August 2018).  Telecoms and large technology companies are leading this charge and mainstream Fortune 1000 companies are not far behind.  In this rush to acquire renewable energy for sustainability and green energy supply, many companies are forgetting that these deals can come with substantial financial exposure.  Google may be able to afford significant losses in its renewable energy portfolio.  Can you?

The effective lifetime of renewable energy deals, such as a power purchase agreements (PPAs), is measured in years.  Increasingly, corporate entities are opting for financially-settled PPAs known as “contracts for differences” (CFDs).  The number of CFD variations in the market is staggering, with each delivering a slightly different risk profile to the buyer.  If the buyer is on the wrong side of this risk when market conditions change, he can end up with many years of making monthly “wrong way” payments to the seller instead of collecting revenue from his PPA.

Obviously, it is impossible to predict the weather, energy prices, availability, and demand over this entire future horizon, though each of these components can materially impact the contract value.  Experienced energy companies, such as large utilities and global energy trading firms have been managing these risks for a long time. How, do they do it?  They use sophisticated simulation models to predict the range of possible future outcomes and the most likely outcome.  This allows the company to develop an expectation of their future cash flows, quantify the uncertainty around future cash flows, and mitigate risk using custom-tailored risk-reduction strategies. 

If the renewable energy seller is using these techniques to construct a renewable PPA for the buyer, you can be sure the deal’s terms are skewed in their favor.  My hope is that this primer will give corporate renewable energy buyers a better understanding of the financial risks involved and how to mitigate them.

Electricity Prices

Electricity (or power) prices exhibit dynamics that are fundamentally different from other financial products or commodities.  One driver of these differences is that it is not cost-effective to store large quantities of electricity, except over short time periods. Another driver is that power system reliability (i.e., “keeping the lights on”) requires a constant balance between production and consumption. At the same time, electricity consumption (demand) depends both on weather and the intensity of business activities and can be highly uncertain. The availability of electric generating assets to meet demand and the associated costs to do so ultimately determine electricity prices at any particular time and location on the grid. These unique and specific characteristics lead to price dynamics not observed in any other market, such as seasonal and intra-day fluctuations that tend to differ widely by geographic location.

Since it is impossible to know with complete certainty what the electricity price will be in any given day/hour in the future, price simulation techniques have evolved to capture the unique dynamics of electric power markets and to simulate future outcomes in a probability-weighted format. Analysts want to know both the range of possible prices and the probability that a particular scenario will occur.  For example, we know that in some electricity markets prices can range between negative values and up to $9000/MWh.  However, it is equally important to understand the probability and timing of large price movements, as well as how prices move in relation to other factors.

MidC_DA_TimeSeries-energy-analytics-cquant

Example of simulated hourly day-ahead electricity prices for the Mid-Columbia trading hub.  This chart shows 5 simulation paths.  Notice the strong seasonal shapes in each of the four seasons, as well as marked differences in price volatility throughout each year.

Robust price simulation models employ a Monte Carlo, mean-reversion, jump-diffusion methodology to simulate electricity prices.  This methodology was first introduced to the financial markets in the 1990s and migrated to the energy markets a few years later.  They have been widely used by traditional energy companies for nearly 20 years.  This robust price simulation methodology allows the financial analyst to have confidence that their contracts are being valued against hundreds of internally consistent future versions of the electricity markets and that the “expected future value” can be relied upon. 

Wind & Solar Generation

Renewable energy is inherently intermittent.  Unlike a natural gas generator that can be ramped up or down at-will, the wind blows and the sun shines based on the irregularities of nature. This makes wind and solar production difficult to predict over long time periods with a high degree of accuracy.  As with electricity prices, methodologies have emerged to simulate future renewable energy production.

The preferred methodology to forecast future renewable generation uses curated databases to access historical wind speed and solar irradiance patterns at any geographic location.  Robust simulation models use these historical weather patterns and the physical characteristics of the renewable generation facilities (e.g. geographic location, wind turbine type and height, solar panel tilt angle, etc.) to generate Monte Carlo simulations of future energy production at the specific asset location.

MthlyGenBoxplot-energy-renwable

“Box-and-whisker” plot of total monthly solar production for the 10 MW Pennsylvania Solar Park. Red horizontal lines denote the expected value (mean), boxes range from the P25 to P75, and whiskers extend to the min and max of total monthly generation, as implied by historical weather data.

The combination of hourly Monte Carlo simulations of renewable energy generation and real-time market prices at the resource node and/or settlement hub provides hourly distributions of market value for the renewable energy produced. The stochastic time series models used to generate the simulations are parameterized from historical market prices and solar irradiance or wind data. The parameterization process ensures the modeled output reflects locational patterns in expected price and renewable generation as well as variability around the mean. As such, simulations capture the real-time interplay between energy production and market prices, allowing users to understand the true value of and uncertainty in energy production at the seasonal, daily, and hourly level for a particular generation facility.

The simulated hourly generation and market prices can also be used to directly value renewable PPAs along with the unique “contract-for-differences” (CFD) financial structures they may entail.  Examples of a few settlement structures that can be directly modeled include fixed price, $0 minimum, collar, revenue share, and many others.

Cash-flow-analytics-energy

Simulated contract-for-differences monthly settlement amounts (year 2022 only) for a PPA struck at $35/MWh for the Pennsylvania Solar Park.

How does this analysis help Renewable PPA buyers?

Some corporate buyers of renewable PPAs have experienced significant “wrong-way payments” and financial losses since contract signing.  In almost every case, those buyers used the analysis provided by the seller as their reference of fair value for the contract.  It comes as no surprise that the sellers often used optimistic forecasts of future value and downplayed the effects of uncertainty.  A robust financial risk assessment and valuation model helps buyers of wind and solar renewable energy understand the true value and risks associated with long-term PPAs. Moreover, it provides an unbiased assessment of fair contract value from an independent third party that has no stake in the deal. This can add significant leverage to the buyer’s position at the negotiating table.

With a proper financial analysis, PPA buyers should be able to report fair market value for existing or proposed PPAs, as well as the degree of uncertainty around monthly settlement cash flows and the effect this uncertainty has on overall contract value. Armed with this information, renewable PPA buyers are in a much better position to negotiate on contract terms and pricing, resulting in higher rates of return over the contract lifetime. Buyers no longer need to rely solely on the seller’s own assessment of value when signing long-term renewable energy contracts.

ReAssure® for Renewable PPA Buyers

cQuant.io’s ReAssure model uses well-established valuation methods to demonstrate fair value and financial exposure for renewable assets and power purchase agreements (PPAs).  The use of Monte Carlo market price simulations coupled with simulations of generator production is a technique that has been used for decades to manage portfolios that include dispatchable thermal generation assets and financial contracts. With ReAssure, cQuant.io has applied these seasoned techniques to the valuation of wind and solar projects and PPAs.

ReAssure combines simulations of hourly electricity prices and hourly renewable energy generation to calculate a broad range of possible future scenarios of the true market value of generation. Each simulation is custom-tailored to project-specific parameters such as precise geographic location, nameplate capacity, tilt angle (solar), turbine manufacturer and model number (wind), and market pricing node or hub.  These simulations capture the real-time interplay between energy production and market prices, providing a comprehensive bottom-up analysis of the true value of and uncertainty in renewable asset or PPA cash flows.

The example output image below shows the expected fair market value of a 5-year renewable PPA, as well as the simulated distribution of possible.  The expected value gives the renewable PPA buyer a single number to consider when making a buying decision while the spread in the distribution provides an indicator of the risk in the deal related to variability in renewable energy generation and wholesale electricity market prices.  With these key metrics in-hand, buyers can make an informed decision to purchase a particular renewable energy contract. They can also quickly compare value and risk across many different proposed contracts and locations to determine the best deal for their renewable energy portfolio.

Simulated market value of electricity generated from 2018 through 2022 for the Pennsylvania Solar Park.

About cQuant.io

cQuant.io is an industry-leading provider of cloud-based energy analytics solutions. cQuant’s advanced analytics platform provides sophisticated energy-focused models across a wide range of industry verticals, from renewable energy project development and contract structuring to thermal generation asset analysis, hedge optimization, and exotic derivative valuation. Visit www.cquant.io for more information.

SCHEDULE YOUR DEMO TODAY

August 29, 2018

Energy Risk Management Software: Build, Buy or Assemble?

When it comes to energy risk management analytics, the classic dilemma that most energy companies face is “build or buy”.  A dilemma is a situation in which a difficult choice must be made between two equally undesirable alternatives.  Later I will introduce the new alternative of assembling a risk solution.  First, let’s spend a moment describing how most energy companies get themselves into the build versus buy dilemma.

A typical risk management life-cycle might look like this:  A young energy company (let’s call them Big Energy) operates with a financial spreadsheet that forecasts revenue, costs and margin.  At some point, this simple view proves insufficient.  So Big Energy hires a risk analyst and/or quantitative modeler to calculate exposure under various scenarios.  The risk analyst creates new tools and happily describes these creations to his/her colleagues.  Soon Big Energy has dozens of independent analytic tools, often in spreadsheets. For the risk analyst, this is a lot of fun!  It’s so much fun that they might not want to openly describe the deficiencies in those analytic tools.  We all know what comes next….

At some point, Big Energy experiences a large unforeseen financial loss and senior management realizes their mistake.  Their risk analyst was not able to calculate or mitigate the portfolio risk from simple in-house models.  So, Big Energy hires a new senior risk manager who promptly puts out an RFP for Energy Risk Management Software.   All the big-name risk vendors apply as well as consultants that act as system selectors and system integrators.  Of course, the consultants are in bed with the big software vendors.  If you pick any of the big system selectors, they will funnel your business to their favorite (massively difficult to install) risk software company.  That is the way the game is played.

Big Energy spends 12-18 months picking a system integrator and energy risk management software vendor. They are about to buy, install and maintain an enterprise risk management solution. Based almost entirely on their ability to spend money, Big Energy is now facing a $2M-$10M expenditure over several years.  Very typically, they will only receive 60%-80% of the functionality they requested.  To fill these gaps, the company continues to buy new services from the software vendor.  Big Energy has no choice; they are essentially stuck.

Like a drunk on a bender, the risk management software vendor happily spends Big Energy’s money until they are cut off.  Eventually, Big Energy does just this, but they still have only received a portion of the functionality they need.  What happens now?  Big Energy’s staff will continue to build solutions outside their risk management software, and new processes will be created to manage multiple independent analytic tools, again.

Raise your hand if you know what I am talking about.  In which stage of this cycle do you find yourself? 

In the energy industry, build or buy are equally undesirable choices.  What if you could access risk models on-demand, with no deployment projects, upgrades or long-term commitments? 

At cQuant.io, we allow our customers to access sophisticated energy models, on demand.  Based on the requirements of each energy portfolio, customers assemble component models into proprietary analytic chains creating a solution just for them.  Energy companies can also access our team of PhD quants to quickly build new analytic models at a fraction of the cost of hiring their own quants.

Energy Risk Management Software Done Right

Nobody likes a dilemma.  The choice to build or buy is a trap on both sides.  In this dilemma, energy companies must decide to either go into the software business (build) or bet the farm (buy) on a single software vendor. cQuant.io is offering a 3rd choice – assemble and start using your risk solution at cQuant.io today.

This article was written by David Leevan. David is the Managing Director of cQuant.io, a SaaS platform for energy analytics

Need a solution for your energy analytics? We can help! Request your free demo today.

Start Your Risk Management Analysis Today

July 16, 2018

Nine Questions with Devon Williams, Risk Manager at Grant County Public Utility District.

Recently, I had the opportunity to sit down with Mr. Williams. In the following interview, he provides sage advice on risk analytics and how to grow awareness of risk-based thinking within an organizational culture.

Q1:  Mr. Williams, as the Risk Manager for Grant County Public Utility District, what are your primary responsibilities?

Chiefly I am tasked with developing our enterprise risk platform and integrating it with company operations.  That includes developing staff both in my group and outside, and sponsoring processes to improve risk-based thinking.  We are focused on doing so right now with our audit function, our capital budgeting and operational budgeting functions. Also, we review deals proposed to hedge our power portfolio and customer rates.

Q2:  Establishing a culture that is aware of risk methodologies and risk-based decision making is difficult.  How are you doing this?

From a high level we are sharing and sponsoring the concepts of probabilistic thinking in the planning and evaluation process both on the front end and back end of our efforts.  Sometimes this happens formally through our job duties.  A lot of it happens informally.  Luckily, we have the confidence of our executives.  We get asked to advise committees and groups or projects as well.

Q3:  Can you provide an example of how this way of thinking impacts a risk management process, at Grant County PUD?

For example, we are doing test evaluations of capital projects.  Instead of using an IRR type of criteria, we are proposing a risk-adjusted return on capital.  That is serving in the discussion long before the project would ever be done.  It serves to flesh out what are the project risks and encourages the discipline on the part of the proposers that know best about it to attempt to quantify those risks and concerns, not only of the project but of its outcomes.  Not all avoided costs are certain but let’s go ahead and put what we think are the avoided costs in the return.  This is actually very beneficial if we are willing to step in the risk space as opposed to keep things on a very engineering economics level.   

Q4:  You seem to be advocating for a more dynamic approach to risk management, rather than a static use of standard risk metrics.

Yes.  There is the concept, just like the outcomes can always change, that so too can the inputs.  It’s a little bit like an agile software release, where we need to continually refresh the information we are getting and the information we are bringing to our meetings in different ways as we learn more and make more decisions.  Keep the discussion at a level that keeps people engaged, because that is how you continuously get better and better information going into your work product. 

Q5:  You mentioned an agile approach.  When you think about analytic models or an analytic process, what do you look for?

I like to be able to have very clear access to the inputs, controls, and settings in the model.  I think that sometimes what gets lost with analytic models, powerful wrenches that they are, is the communication of the sensitivities and central assumptions that are governing the work product.  The better we do that and the better we teach other people to use what we created, our work becomes more usable and accessible by everyone across the business and people are a lot more willing to apply it.  So, instead of a dashboard, have a control panel on the model.  For example, we are coming up with a different rate for a different class of customers and my group was asked to develop some risk-based components for that.  We came up with a model that is more of an analytic framework.  It explains the inputs and key assumptions, and suggests how these need to be refined.  We highlighted, placed comments and explanations of what these inputs and assumptions are and how they could be applied.  This really builds our engagement and greatly enhances the likelihood that the work product will be used. Therefore, you have a chance to impact the company culture.     

Q6:  These are important lessons for communication and utilization.  When talking to senior management about risk, how do you communicate risk in a way that is actionable? 

I think it’s a two-step process. First this to share some conclusions in plain English, such as “our analysis is indicating X & Y and it is based in these sensitivities and root causes”.  And then dig into the foundation of why we think that.  Some more technical calculations may come up but having the context of what conclusions we are drawing from it makes its easier to understand the calculations and appreciate them. I like to focus on the simple first and get into the underlying complexities after we have engagement at the higher level, which is the reason we are there.

Q7:  Do you make use of risk metrics like VaR or CFaR/GMaR, and how do they influence concrete actions? 

There is a project where we have done a portfolio VaR and a VaR stress that turns into a cash-flow test.  At no point in the discussion, writeup, conception did I use the words VaR or cash-flow-at-risk.  Those are words that mean plenty to me, but they don’t necessarily mean much to many of our stakeholders.  I see it as part of my job and my team’s job to demonstrate interpretations of the results and share the specifics of how those results were achieved in a later discussion.   I think that is probably the key.  If they are presented as concrete concepts upfront, you have a chance to getting to concrete actions.  If you weigh people down with technical concepts up front, they are likely to disengage.  Its on me to make my work product work for everyone else.  That makes me no different than a chef or a contractor.  We all have to please our customers.

Q8:  Earlier you mentioned improving access to capital.  What are primary lessons that you have learned to make sure access to capital is maintained?

Investors can and do read financial statements, but I think getting that quantification of potential underlying outcomes, and what likely and concerning events could cost, having them understood and having the value of their limits understood is probably the greatest challenge to improving access to capital for the company.  Explaining to people why we might impair a deal’s value or a forward projection with probabilistic outcomes can be difficult, but it helps the over-all product considerably. 

Nothing is really Tab-A into Slot-B in our business.  There are a lot of things that can go right or wrong or different.  But having an appreciation for the likely path and some potential surprises serves to make a more understandable operating envelope.  By including the exposure to both the upside and downside our projections are more believable and this improves access to capital. 

Q9:  Finally, what is a relatively new risk factor that has emerged for you in the last few years?

While not entirely new, one would be the claims and litigation culture.  This continues to build, and the degree and weight are greater than what we would have thought 10 years ago.  Consistently low gas volatility is another one.  That is a new world for electric utilities and their cost models.  I would also add, in a way that is surprising people with the internet of things, the costs and security of tech integration is a major opportunity and risk for utilities that we continue to monitor.

This interview was conducted by David Leevan. David is the Managing Director of cQuant.io, a SaaS platform for energy analytics. 

Need a solution for your energy analytics? We can help! Request your free demo today.

SCHEDULE YOUR DEMO TODAY

April 24, 2018

Moneyball for Energy Portfolio Analytics

Energy portfolio managers have an on-going dilemma.  When it comes to market and portfolio analytics, when is your analysis good enough?  Unfortunately, financial constraints often drive this decision.  I hear this all the time, “My management will not allow us to add another expensive solution, so we are doing that work on a spreadsheet.”

Does this describe your situation?

  • Energy portfolio is increasing in size or complexity
  • Current analytic tools do not capture the portfolio’s full value or risk
  • Need better portfolio analytic tools, but
  • Cannot add more staff
  • Cannot spend more money

Conclusion: I cannot improve my portfolio analytics.

Can the principals of Moneyball help?

Moneyball shot to attention in 2003 with Michael Lewis’s book “Moneyball: The Art of Winning an Unfair Game”, telling the tale of Oakland Athletics baseball team and their incredible success against the odds, spearheaded by general manager Billy Beane.  Beane took a team on a limited budget to an American League record of 20 consecutive wins, on the way to finishing top of the American League West in 2002.

The basic principal of Moneyball is to correctly value your own assets and other available assets in the market to construct the best outcome on a limited budget.

Like baseball, the energy market is an uneven landscape with big, rich companies and smaller resource constrained companies. Managers of smaller energy companies need to find better analytics to compete with larger energy companies.

The traditional methods to improving energy portfolio analytics are to (1) hire your own Ph.D. quants, then build and maintain in-house tools, or (2) hire expensive consultants for analytic projects, or (3) buy expensive ‘enterprise’ software. These options require significant upfront costs, time, and a lot of optimism that it will work.

The most popular alternative to the options above is to maintain a simple spreadsheet and hope that you do not make a mistake.  The problem is that eventually you probably will overlook something important, and other market participants will be there to take advantage of you.  In fact, your counterparties may be taking advantage of you today.

cQuant.io offers the energy manager a new option to consider.  Our sophisticated analytics are available on-demand in our web application.  We offer analytic models for thermal & renewable assets, storage, load, market analysis, hedge valuation and risk management.  There is no waiting, no IT projects, and no hidden fees. 

That’s why we like to say:  cQuant.io eliminates 100% of the waiting, 100% of the risk and 80% of the cost of adding sophisticated analytics into your management process. 

Maybe it’s time to add more power to your portfolio analytics.  cQuant.io can help you to start winning big today.  Want to see examples of customer driven analysis?  Read our Use Cases.

David Leevan is the Managing Director of cQuant.io, a SaaS platform for energy analytics.    

Are you ready to improve your energy portfolio analytics? Request your free demo today.

SCHEDULE YOUR DEMO TODAY

March 6, 2018

Nine Questions with Ivo Steijn, Head of Model Risk Management at Silicon Valley Bank

This past week, I had the opportunity to sit down with Dr. Ivo Steijn. I have known Dr. Steijn for several years and always found his perspectives to be invaluable. In the following interview, he provides sage advice on energy analytics, model validation and how to see major market changes coming.

Dr. Steijn, you have had an impressive career in quantitative analysis & risk management for several firms in both Energy and Finance.  You have worked in senior roles at TXU, Southern California Edison, State Street and now most recently at Silicon Valley Bank (SVB).  Can you tell us what you are doing at SVB?

Sure, I head the Model Risk Management department.  We are responsible for the validation of all quantitative models in the company, together with all of the administrative superstructure that goes with that: our model inventory, change logs, etc.

What is your view of the relative sophistication in portfolio analytics and risk management between the energy & finance sectors?

I think the Finance world is a little ahead.  They have got the whole portfolio approach to risk management served up to them in the 1950’s by Markowitz (https://en.wikipedia.org/wiki/Harry_Markowitz) and it has been a standard paradigm there ever since. 

For the Energy industry it’s a lot newer.  I also think in the financial world, they throw more money at the problem.  They develop more systems; they hire more developers.  In the energy world its still fairly new.  There is not a lot of really good software, although that is changing over the last couple of years.  So, energy is a little behind finance.   

Do you think there are untapped competitive advantages in the energy sector for companies that want to do analytics better?

Oh absolutely!  In the first place, there are not a lot of good choices in the energy world.  That is strike number one.  In the second place, the choices that are available are fairly inflexible.  You buy a giant software package, and then you spend weeks installing it, more weeks to customize it and then it might not give you what you want.  There is certainly a need for a solution that is more tailored to a wide variety of customers, that people can use in different ways, that provides people with more options. 

What is a common oversight that you encounter in businesses which use quantitative models? 

The oversight that I most commonly encounter is thinking that once your model is okay, there is no model risk.  A lot of our work consists of hunting down bad models and fixing flaws in them.  But a major oversight is that even after the flaws are fixed, the model still has significant uncertainty around it.  It may contain estimated parameters, which is a best guess at an uncertain number. 

Your model is not something that is handed down to you on stone tablets on a mountain top.  It is a result of a messy statistical process which gives a fuzzy result at the end of it.  This has consequences for how we should be looking at models and using the output.      

Can you give us an example of model uncertainty?

My favorite example.  Let’s say you have a standard two commodity portfolio of power & gas.  Your Monte Carlo price simulation process uses a correlation structure between power prices & gas prices.   The correlation structure is calibrated from historical data or from options.  That correlation parameter, used by the model, is often viewed from this point forward as a constant of nature.  It is not!  There is a lot of uncertainty around it.  Once you start taking the uncertainty of critical model parameters into account, the results of your models can change. 

How should analysts address model uncertainty? 

The easiest and fastest way is to play with sensitivity analysis by wiggling the parameters around to see if the model results change significantly.  This is easy to do, and we think that sensitivity analysis is something that everyone should do for every model.   The gold star solution is to start your Monte Carlo approach by doing Monte Carlo analysis on the model parameters themselves.  Generate a Monte Carlo parameter set and do your price simulations based on those simulated parameters.  Repeat this for each simulation. This is a much more comprehensive approach.

Should companies be worried about structural changes in energy markets? 

If they are not worried about it, they are not reading the newspapers.  Anyone working in energy, particularly in CA where I worked for a long time, knows that gas fired power plants used to be the marginal asset.  You worried about gas and power prices, and that was it.  These days, for many hours each day, renewables are the marginal asset.  That completely changes your price process.

This structural change snuck into our portfolios gradually.  Other structural changes can come up faster, but this one was more gradual and it is spreading to the rest of the country.  This is a major regime change that has consequences. You basically have to redo your whole portfolio analysis.   

How can we see these regime changes coming?

The standard truth about regime changes is that nobody sees them coming.  I think that, for some regime changes, you actually can see them coming, you just don’t know when they will arrive.  My advice is to do your modeling work in advance.  Model the new world where renewable power is the marginal asset and then play with different scenarios around when the regime change occurs.  That will give you a more realistic view of what the long-term outlook for your portfolio may be.  You don’t know if it’s coming this year or the next, but you know it’s going to happen.  Well, start running your scenario analysis now. 

Finally, what is your view of how technology changes such as cloud computing is affecting the analytics landscape?

The old paradigm of a giant piece of software sitting somewhere in your building is changing.  Increasingly, we are working with very thin client solutions and accessing all of our computational machinery via the cloud.  Why would you have all that computational machinery at your company?  It’s just not efficient.  I don’t see lot of this kind of development happening yet, but it’s happening more and more.  The old paradigm of locally installed software, I think that is going away.   

Ivo Steijn is Senior Director, Model Risk Management for Silicon Valley Bank where he is responsible for all model validations and chairs the Model Risk Management Committee. Prior to joining Silicon Valley Bank he was a VP in Model Risk Management at State Street in Boston, and he headed the Model Validation department at Southern California Edison for 12 years. He holds a MA and PhD in Econometrics from the Free University in Amsterdam, the Netherlands.

This interview was conducted by David Leevan. David is the Managing Director of cQuant.io, a SaaS platform for energy analytics. 

January 29, 2018

Energy Risk Management: Five Steps to Improve Your Process

Choose the right risk metric.

Energy portfolios vary greatly from company to company and location to location.  No two energy portfolios are the same, and risk management strategies must be tailored to the unique risk factors of each portfolio.

Many companies use a traditional Value-at-Risk (VaR) metric to report risk to their Board and to highlight risk to the portfolio in quarterly financial reports.   But what action does your organization actually take if your VaR goes up by 25% in one week?  If you don’t have a great answer, VaR may not be right for your portfolio.  At cQuant.io, we use both VaR and Cash Flow at Risk (CFaR) methodologies to assess and report portfolio risk, depending on the type of portfolio and the goal of the risk reporting.  Here is a simple comparison of the two:

Which is more appropriate, VaR or CFaR?

  • VaR is a risk metric that provides not only the value the portfolio could lose over a given time interval in a “worst case scenario”, but also a measure of the statistical confidence around that estimate. For example, computing the VaR for a large portfolio of financial contracts may suggest there is a 95% chance that portfolio will not lose more than $1M over the next 5-day period.  VaR is most appropriate for assessing risk over short time periods (less than one month), particularly for positions that can be unwound in a few days.
  • CFaR is a more appropriate risk metric for companies with physical assets, customer load or complex structured transactions.  These types of portfolio elements cannot typically be sold or significantly modified in response to short-term market moves. CFaR provides a detailed analysis of cash flows through time based on many simulations of possible future states (more on simulations in item #3). The result is a distribution of future value and costs for any time bucket (day, week, month or year) on any portfolio item (gen asset, storage, customer, deal, commodity, counter-party), or for the portfolio as a whole.  The results can be proactively used to reduce risk, lower cost and improve portfolio performance.
  • Why use both?  Many energy portfolios contain both long-term and short-term assets.  Using both CFaR and VaR can provide advantages for more diverse portfolios.  Additionally, more sophisticated VaR models allow users to slice through portfolio dimensionality to report net position by commodity, value at risk by trader or counter-party, and other useful metrics that can be used to surgically target and mitigate risk from specific portfolio components.  Taken together, combined VaR and CFaR analysis can enhance active risk and portfolio management.

Choose the right risk factors.

What is driving the risk to your particular portfolio?  Price volatility, weather uncertainty, basis risk, customer migration, congestion, government policy or regulation?  Understanding the underlying “risk factors” in your portfolio is crucial to modeling your risk appropriately.  As your portfolio responds to changes in energy markets and supply/demand dynamics, even the sharpest intuition can mis-identify the riskiest portfolio elements.  A good risk management process can help keep the focus on what poses the greatest threat to your company financials.

Choose a Great Simulation Model.

It seems that everyone has a simulation model these days.  At the simplest level, these are Excel plug-ins, closed form models or a bunch of legacy code from some analyst that used to work at your company.  Here is the danger – a poor simulation engine can be much worse than having nothing at all.  At least when you have nothing, you know it.  A poor model may add uncertainty in a way that is not market-consistent, may misrepresent important relationships between related commodities (e.g., correlation between power and gas prices), or may simply drift out of calibration as market conditions evolve.  Inaccurate simulations can result in actions that are completely inappropriate for your real risks.  Its important to validate any simulation model against actual market data regularly to ensure it’s doing its job.

Include everything in your model. 

Too many companies get lazy with this.  “My model includes about 90% of our portfolio, but the more complex transactions we value in a separate process.”  It may be the case that those exotic transactions are driving an outsized portion of your risk!  Recognize that big risks can come in small packages and allocate staff (or vendor) time to include all transactions into your model to get a comprehensive and accurate risk analysis.

Turn results into actions. 

Surprisingly, this is the most common gap.  Perhaps you have done all the hard work to set up a comprehensive risk management system, but the results still aren’t affecting real portfolio decisions.  Try holding monthly risk committee meetings that include your financial, trade and asset managers along with at least one C-suite sponsor.  The volatility in energy markets is the risk manager’s best friend.  The next Polar Vortex or Bomb Cyclone is lurking just around the corner.  Eventually, you’ll be able to say, “I told you so”, and win the day.

Was this article helpful?  Are there additional subjects that you want us to write about?

Click here to leave feedback.

Need a solution for your energy analytics? We can help! Request your free demo today.

Start Your Risk Management Analysis Today

This article was written by David Leevan. David is the Managing Director of cQuant.io, a SaaS platform for energy analytics. 

January 18, 2018

Energy Analytics Revolution is Here!

I have been writing recently about the risks associated with buying enterprise software, advances in cloud computing, and the benefits of software-as-a-service. I have tried to keep those articles fairly vendor agnostic. This article is going to be more self-serving. I am basically going to describe why we at cQuant.io decided to build a new type of energy analytics company.

What is wrong with the current choices? How can the energy industry benefit from something new?

As many of you know, energy companies are driving their business with analytic decision making. The best energy companies spend $millions on analytic solutions and are constantly looking for ways to improve. Smaller organizations often outsource analytic decisions to outside firms because they cannot afford to spend much on analytic solutions and analytic staff.

 Why is this happening? Is there another way?

Let us first examine the current options for adding or improving your analytics. There are basically three choices that you have as a customer:

  1. Build it yourself – this requires the customer to go into the software business. The company must hire staff to build the quantitative models, the UI, the database, the system integrations, etc. Then maintain and improve it over time.
  2. Hire a consultancy – this is outsourcing the software development. Now you must pay consulting fees for the life of the project, which is often quoted in months but requires years to complete (funny how that works).
  3. Buy a vendor solution – analytic software vendors normally have solutions that are both enterprise ready and highly configurable. Unfortunately, they also require large upfront fees, deployment projects, and annual maintenance contracts.

These options require the buyer to shoulder large upfront risk. Push your money in the middle of the pot and hope that it all works out. To address this risk, many energy companies employ bureaucratic procurement processes.

It can take years to get the solution in place.

We decided that the energy industry needed another choice. Our goal was to eliminate 100% of the risk, 100% of the waiting and over 80% of the cost associated with the existing choices.  So, we began working with top quantitative experts to build sophisticated energy models and we deploy those models in a modern cloud computing platform.

Now energy analysts have a place to go to find and use analytic models – on demand! Our models can be used for a day, a week, a month, or a year. We are adding new models every few months.

In truth, we are just at the beginning of our journey. Our team has tremendous plans to revolutionize the energy analytic landscape. You deserve more and better choices. We hope to be a part of this revolution.

This is just the beginning and I hope that you will join us! 

This article was written by David Leevan. David is the Managing Director of cQuant.io, a SaaS platform for energy analytics. 

Request Your Demo Today

Scroll to top