The Kathmandu Terai/Madhesh Fast Track Road – A possible financial disaster in the making and need for further critical analysis

Recently there has been a lot of activity with regards to Kathmandu Terai/Madhesh Fast Track Road (KTMFT).  DPR was awarded to IL&FS consortium in March and now there is talk in the media for the same consortium to take on the project on a BOT basis which part of the growing popularity of the PPP (public private partnership) movement in the award of infrastructure projects.  Any PPP is a sharing of risk and return between the private and the public parties and this sharing should be based on who is able to take on the risk properly and how the cost of those risks are borne by the parties.  The distribution of risk and return should be balanced between the parties.

With what little information that is out in the public domain, one gets the feeling that in the KTMFT case proper due diligence and balancing of risk and return on the project between the parties have not been done.  There also appears to be a fundamental flaw in the overall analysis upon which certain terms and conditions have been defined for the project.  I just want to focus on two issues that seems to make this project very one-sided and to the detriment of the Nepali taxpayers.

Table 1: Minimum PCU and MRG as per LOI to award dated Feb 6, 2015

The LOI (http://www.mopit.gov.np/files/download/LOI%20Notice.pdf) intent published in the MOPIT (Minsitry of Physical Infrastructure and Transport) web site mentions that it expects to award the IL&FS consortium the project based on the numbers they have provided with respect to the cumulative passenger carrying units (or vehicles) of 558,356,865 and the present value of the minimum revenue guarantee (MRG) of NPR 317,040,820,961.09 over the 25 year operations period.

First let us dissect what the cumulative PCU implies.  Essentially, GON will be guaranteeing that the concessionaire will get 559,365,865 million vehicles using the road over the 25 year concession period.  This translates into 22,334,275 (559,365,865 / 25) vehicles per year or 61,190 vehicles per day (22,334,275 / 365) or 2,550 (61,190 / 24) per hour or 43 per minute (2,550/60) over the entire operations period.  Anybody who has driven from Kathmandu to Mugling (the busiest section of Nepali highways) knows that is nowhere the case and it will take decades to reach that level.  An article in Kantipur on 2014-03-11 states that a survey revised down the expected traffic on this road to 3,000 per day from the ADB estimate of 7,000 per day.  (http://www.ekantipur.com/2014/03/11/business/study-downgrades-expected-traffic-on-planned-fast-track/386571.html)

Table 2: MRG based on Minimum PCU and impact of 10% shortfall in Minimum PCU

The minimum revenue guarantee is based on this figure.  If you assume monthly payments over a 25 year period, to get the minimum revenue from the implied PCU, the toll fees have to average 1,550 per vehicle or 20 rupees per kilometer. (In Excel use the PV function with the rate set as 10%/12, Term as 25 x 12, PMT as 22,334,275/12 x 1,550.  You will get a PV very close to the PV of given MRG.)  In India for cars and jeeps, toll fees are usually less than 1 Indian Rupee per kilometer.  There is even talk about exempting the fees to private cars and jeeps. (http://www.hindustantimes.com/india-news/highway-toll-may-be-exempt-for-private-vehicles-levies-kick-in/article1-1298953.aspx)  The implied toll fees in the MRG will make this one of the most expensive toll fees in the world.  Therefore the willingness to pay for the use of the toll road by non-commercial users may not be there. According to media reports, toll fees for KTMFT road is expected to be 800 rupees for two wheelers and 1,600 rupees for cars and jeeps. For a person earning 30,000 rupees per month a two-wheeler round-trip on the KTMFT will cost him 5.2% of his monthly salary

Add to this, the fact that an alternative route to the Terai, the Kathmandu Hetauda Tunnel Road, has also been awarded to a different group of investors.  If this gets off the ground and completed, most of the container/commercial traffic from Birgunj making the trip to Kathmandu will prefer this shorter route which means that the average implied daily PCU will be even more difficult to achieve. Planners should also realize that for traffic to and from the west and far west, both of these roads do not necessarily lead to significant savings in fuel costs and only marginal savings in terms of time.  Existing highways are being broadened which means that the diversion of central and east bound traffic to these planned toll roads will make them less congested.  West and far west bound traffic might find that the benefit of using the toll roads are not as compelling for them.

It is therefore very certain at the outset that the Nepali taxpayers will be paying IL&FS consortium compensation fees as a result of this MRG.  We will be entering this agreement knowing very well that the MRG will start clicking from the very first year and most likely will continue for the entire 25 years.  A 10% missed target over the 25 years implies a pay out to the concessionaire of 31.7 billion rupees or approximately 317 million USD at todays current FX rates in present value terms.  Clearly the risk and return associated with this project has not been thought through properly. 

It appears that the GON has a clause that assures that it will get 80% of revenue for vehicle traffic above the PCU.  But if the minimum PCU is itself unachievable this benefit-sharing term is meaningless.

The second question I want to tackle is the value transferred by the subsidized financing provided by the government.  The government has two subsidized financing schemes.  A grant not exceeding 15 billion rupees.  If this is part of the equity and the government is going to share in the profits proportionately, I may have very little to say about it.  But if this is a grant and the concessionaire does not have to pay it back with the relevant return then this is 15 billion of value transferred to the concessionaire from the Nepali taxpayers so obviously not in Nepal’s interest.

The other subsidized financing is with regards to 75 billion rupees in long term low interest loans at 3%.  In an article in Kantipur IL&FS consortium says that it could build the road with a commercial loan at 13%.  (http://www.ekantipur.com/2015/07/22/business/hopeful-fast-track-builder-asks-for-help-with-funding/408256.html) Let us use this to determine the value GON is transferring to the concessionaire on the 75 billion at these rates.  Let us assume that payment will be made in equal quarterly installments.  So all you have to do is determine the quarterly payment under the two different interest rates and then calculate the present value of the different in payment.   

Table 3: Estimate of value of interest rate savings given as per LOI.
Payments are equal quarterly payments so Loan is paid off at the end of 25 years.

Let us use the 10% that MOPIT mentions in its various documents when calling for RFP for the project to calculate the present value.  How they came up with this number is in itself a big question but we will not discuss that here.  The difference in payments will be 1.47 billion per quarter whose PV will be 53.91 billion or 72% of the face value of the loan. We have not taken into account the grossing up of interest payments that are done during the construction period.  This is in essence the value that GON is transferring to the concessionaire relative to the market.  The question is what is the GON or Nepali taxpayers getting in return?   

The above analysis is based on the fact that the loan will be in NPR.  If the loan is in USD, then 3% rate is just 0.14% over the 30 year USD treasury rate.  Obviously this does not reflect the risk of the project.  Similarly, let us hope that this is not a fixed rate loan because it is evidently clear that interest rates in the US will be rising.  Let us also hope that the government has not provided an FX rate guarantee. This has its own implications.

From what I have read in the media and the information that is available on MOPIT’s website, I find that the anticipated award to IL&FS consortium is very one sided primarily to the benefit of the concessionaire.  The concessionaire is not even bringing financing to the deal but instead is getting significant subsidized financing and also ensured a minimum return that is not feasible at all from the anticipated vehicle flow dynamics.  I have no arguments with the PPP model as long as proper due diligence has been done and we have taken balanced approach to the distribution of risk and return between the parties involved. 

As is, this project is a disaster in the making and going to be an embarrassment to the GON and the Nepalese taxpayers in the future.  As a concerned Nepali citizen, I would like the responsible authorities to undertake a proper analysis to ensure that the award of this contract is done only after ensuring a proper risk return analysis to all stake holders involved.  

Finally if the government is going to arrange all the financing then why not use the Build, Operate and Maintain model where the concessionaire gets a fixed percentage of the total construction cost as fee during the construction phase and a fixed percentage of annual gross profit during the operation and maintenance phase.

It is clear that GON/MOPIT has not done a proper due diligence regarding the risk and returns associated with this project. They should either get the IBN (Investment Board of Nepal) or an independent financial team to review the terms and conditions and ensure that all parties share in the risk and return in a proportionate manner before the contract is actually awarded.

Note:  I have also read in some media reports that the operation period will be for 30 years and certain terms and conditions with regards to PCU and MRG have changed.  It is therefore possible that certain issues I have mentioned in this blog may not reflect the current status of the project.  Regardless, it is very important to take stock of how realistic the assumptions going into the analysis are since they will drive the risk and returns of the project which will be shared by the parties involved, in this case GON and the IL&FS consortium.

What’s in a USD PPA (Power Purchasing Agreement)?

Recently, I came across two articles one in the Kathmandu Post (KP) and the other in Republica.  The KP article stated that NEA (Nepal Electricity Authority) was amenable to US dollar (USD) PPA in certain conditions and made note of the claims by Nepalese investors in the sector that USD PPAs were essential for attracting foreign investors.  The Republica article mentioned that NRB Governor Dr. Khatiwada was against USD PPA for export oriented projects with investment from non-Indian third countries.  He made the argument that the current level of reserves made USD PPAs for such projects unsustainable and went as far as to say that it could lead to the devaluation of our currency with India, a veiled reference to that the fact the Indian currency peg is currently being supported by the sale of US dollars which could be quickly depleted by these USD PPA obligations if they mushroom. This is not the first time USD PPAs have been in the news and nor will this be the last time.  But what is it, in a dollar PPA, which makes one side claim that it is necessary and the other side that it is unsustainable for the country?

Put simply it is all about the management of foreign currency risk a project is exposed to.  Whenever we talk about a power project today, we are taking about five stakeholders: the government, the equity investor, the debt investor, NEA currently the sole purchaser for electricity for domestic consumption, and finally the consumer.  Foreign investors can come into the equation either as equity investors or debt investors.  Similarly we can have foreign consumers, primarily India, for export oriented projects. A recent article in the Himalayan Times also suggest that Bangladesh is also keen to purchase electricity from Nepal.

Let us begin by looking at several ownership structures and the relevance of USD PPAs with respect to them.

Case 1: Fully domestically owned, domestically financed project supplying the domestic market

The only FX risk these projects have is when they purchase capital instruments or have service contracts for the construction of the projects with foreigners.  The market risk exist only during the construction phase and until the contract terms are honored via payment of necessary amounts.  There is a convertibility risk, that is dollar funds may not be provided by the central bank when it is needed but this risk has very little to do with currency denomination of the PPA.  It should therefore be very clear that these type of projects do not require a USD PPA.

Case 2: Projects with domestic and foreign equity investors, domestically financed and supplying the domestic market only

 From an accounting perspective, these project faces the same risk as projects in Case 1 since the equity investment by the foreign investor will be converted to Nepalese rupees at time of the set-up of the company and each time the equity is called.  The foreign equity stakeholder however does face FX risk on its equity investment and any dividends that it might receive.  However, this investment is no different from the large portfolio investments that funds make in foreign companies by purchasing equity in the local stock markets.  The risk of translation of exited investments and any dividends received during the holding period of the investment is all borne by the investor.  So if a foreign equity investor wants to come to Nepal to invest in a hydro power project that sells its electricity in the domestic market, then while the investor should be comfortable that the investment made and dividends received can be converted to a globally traded currency and repatriated, the investor should also be willing to take all the business and operating risk of running the firm in Nepal which includes taking the risk that his dividends and investments while positive in Nepalese terms may be less in the home currency terms of the investor at time of repatriation.  Otherwise Coca Cola should be demanding that it be able to sell its product in USD terms.  

However this comparison with Coca Cola is not fully correct since Coca Cola can change its price structure to the end consumer to reflect its changing cost structure, while a hydropower project cannot unless it has captive buyers and can operate independently.    Despite the differences, it still makes little sense to have USD PPAs for these type of projects.  Somewhere I read a comment what the impact would have been to foreign investors by the devaluation of Nepalese Rupee from 70 to almost 100 now had they invested.  We should note that NPR devaluation is as a result of Indian Rupee (INR) devaluation and all foreign equity investors who had FDI in India or bought stocks of Indian firms in the Indian stock exchanges were impacted by the devaluation.  We should also mention what if the currency had appreciated from 70 to 50.  Who would have benefited then?  Equity investors have to take the movement of FX rates in stride since that is part and parcel of operating a business in foreign countries.

Case 3: Projects with foreign and domestic debt investors supplying the domestic market only.

It is highly unlikely that a foreign debt investor (except probably the World Bank that will be issuing a Nepalese denominated debt) in a Nepalese hydro power project will invest in local currency debt issued by any project.  If they do invest, they will insist that it be denominated either in their home currency or a globally tradable currency such as the USD.

Unlike equity invested in foreign currency which appears on a projects’ balance sheet in local currency terms, foreign debt if permitted by national government, appears on the balance sheet of projects in the currency of the debt. For simplicity let us say that this is in USD. Whenever there is USD debt on the balance sheet, the project’s outstanding debt obligation in local currency will fluctuate depending on the level of the FX rate.  Similarly, the firms interest payments and possible principal repayment in terms of NPR will also fluctuate. This market risk to FX, as opposed to convertibility risk that all foreign investments are subject to, will exist as long as the debt in USD are outstanding.  Once completely paid, the project is no longer subject to the FX market risk. Obviously the equity investors remain and their investment will be subject to both market risk and convertibility risk but they will be no different from the risk that foreign investors were subject to in Case 2. 

This suggest some boundary rules in the construction of any PPA agreement with foreign investment.  If the objective of the USD PPA is to ensure that projects are able to pay the interest and principal repayment obligation of their foreign currency debt, then any pricing arrangement should be a weighted component of local and USD PPA price.  For lack of a better measure this could be the foreign debt to total financing (debt + equity) ratio.  The applicable term of any USD component of the pricing agreement should not exceed the maturity of the longest USD debt, that is, after the USD debt matures all prices of electricity should be in local terms.  Similarly, the weight of the USD component of the pricing should decline as the outstanding level of foreign debt declines.  We will look at an example on how this might work later.

Case 4: Projects with foreign or domestic equity and debt investment with no sale to domestic consumers

Since the electricity is not being generated for local consumption, a USD or NPR PPA is moot in this situation. The government definitely has a role to play through the signing of a power development agreement to facilitate the trade of electricity across borders and encourage such investments since they will be the direct beneficiary of tax and royalty receipts which can be invested for the larger good of the country.  IPPs who want to sell their electricity to third countries will actually have to sign a transmission agreement which will require them to pay for the usage of the transmission infrastructure to get their electricity to the targeted countries.  Should the domestic market require electricity from such projects, then domestic distributors may enter into supply agreements with such producers in mutually agreed terms which could be in local currency or USD if permitted by the country’s regulations.  Assuming that these will be short term in nature, the impact of such agreements whether in local or USD terms would not be the same as signing of long term USD PPAs.

Case 5: Projects with foreign or domestic debt investment with partial sale to domestic consumers

Case 3 analysis would be applied here but only for the portion that is targeted to domestic investors.  Ad hoc extra purchases can be made based on contractual terms that is not part of the PPA agreement targeted for domestic consumption as mentioned in Case 4.

How could we structure a USD PPA based on the analysis presented in Case 3? 

The first step would be to calculate what the PPA price per unit of electricity would be under two scenarios: In scenario 1, the PPA would be entirely in local terms for the duration of the concession.  In Scenario 2, the PPA would be entirely in USD terms for the duration of the concession.  

Figure 1: Deconstructing a USD PPA

We should keep in mind that when an IPP asks for a USD price for its electricity, then in effect the purchaser (NEA) is giving the IPP an FX guarantee. The IPP could get PPA in Nepalese rupee and then buy an FX contract simultaneously to sell NPR rupees for USD in the market.  Thus, the USD PPA can be thought of as two contracts (see Figure 1): Contract 1 where NEA agrees to purchase electricity in NPR and Contract 2 where NEA agrees to repurchase the NPR from the IPP based on Contract 1 and provide USD in return.  The second contract is a long-term FX contract which has cost associated with it and NEA should charge the IPP for providing it.   Let us for simplicity assume that the local currency PPA is 8 rupees per unit.  Then in current USD terms it would be 8 cents on the dollar assuming 1 USD = 100 rupees.  Now let us assume that the cost of providing the embedded FX contract is 3 cents per unit, then an equivalent USD PPA would be 5 cents per unit.  Please note that there are methods to calculate what the cost of FX contract would be using the relationship between NPR and INR and INR and USD but this is beyond the scope of this article.

Next let us determine what the foreign debt to total financing ratio is.  Let us say that equity is 30%, and that debt is equally divided between foreign and domestic.  So the foreign debt ratio is 35%. 

Now we have all the components to structure a simple PPA agreement that incorporates the above scenario.  If payments between NEA and IPP are done monthly, then we an essentially have a clause that says that of the monthly output purchased by NEA, 35% would be at paid in USD at 5 cents per unit and the remaining 65% at 8 rupees per unit.  USD debt to total outstanding financing ratio would be revised at the beginning of each fiscal year. Price escalation clauses in the PPA would be applied to both components. This will also ensure that once the USD debt component is paid, the PPA price will only be in local currency terms.

Table 1 provides a formulaic view and a worked out example.

Table 1: Formulaic and worked out example of how a PPA could be structured

Assume
Qym	Amount of electricity sold by IPP to NEA in month m in year y in kilowatt hours.	10.4 million KWh
P($,ym)	Price per unit of electricity in $ for month m in year y.	0.05 cents (starting year)
P(N,ym)	Price per unit of electricity in Rupees for month m in year y.	8 NPR (starting year)
D($,y)	debt in $ at beginning of year y.	17,500,000 USD (staring year)
D(N,y)	Debt in NPR at beginning of year y.	1,750,000,000 NPR (starting year)
E(N,y)	Equity in NPR at beginning of year y)	1,500,000,000 NPR (starting year)
X($N,y)	exchange rate for one USD in NPR at beginning of year y.	Assume 1 USD = 100 NPR (starting year)
R($,y)	Ratio of USD debt to total financing = D($,y) * X($N,y) / [ D($,y) * X($N,y) + D(N,y) + E(N,y)]	= 17,500,000 x 100 / (17,500,000 x 100 + 1,750,000,000 + 1,500,000,000) = 35%
USD Payment	R($,y) * Qym * P($,ym)	0.35 x 10,400,000 x 0.05 = 182,000 USD
NPR Payment	[1 – R($,y)] * Qym * P(N,ym)	(1-0.35) x 10,400,000 x 8 = 54,080,000 NPR

Note: the worked out example is based on a 25MW plant that costs 2 million USD per MW to build.  This would require 50 million USD to build implying 15 million USD or 1.5 billion NPR in equity investment, 17.5 million USD in foreign debt, and 1.75 billion NPR in local debt.  Turbine operates 24 hours 365 days with an average efficiency of 57%.  These assumptions may not reflect reality and are being used just for exposition purpose.

How NEA could hedge its FX exposure from USD PPA?

Now that we have looked at how NEA could structure a PPA with the IPPs, let us also look at how it can manage the fluctuations that will result as a result of a part of its expense stream being in USD.  First of all, the NPR is not a globally traded currency and so there is no real market mechanism for NEA to hedge its risk.  However there does appear to be an active USD and Indian rupee forward market.  Given that NPR is pegged to the US dollar, NRB regulations permitting, NEA could use the INR market to hedge its foreign currency exposure as a result of the USD PPA on a short term, say one year, rolling basis. 

Every year as part of its planning process, the NEA could estimate how much payments in USD it will have to make with respect to all of its obligations to IPPs.  It could then enter into hedge contracts with Indian banks or even Nepalese banks (regulations permitting) to sell INR to get USD.  There is a big assumption here that NRB will give INR against NPR for NEA to sell to its FX counterparties to honor its obligations).  Any excess cost as a result of the hedge would then be rolled out to its consumers on a proportionate basis with a revision in prices at the beginning of each year.  If NRB is unwilling, then NEA can still do the exercise of what its cost to put in a one year hedge on its USD commitments and could roll out the anticipated excess cost with a price revision. In this case NEA will bear the cost or benefit of the price revision, i.e. if the dollar appreciates more than anticipated, NEA will have a loss even with the price revision but if dollar rates appreciate less than anticipated, the NEA benefits with the price revision for the given year.

Obviously, this also anticipates that there is political will to deregulate electricity price control.  If the government wants to control the price of electricity and provide consumers with a floor on electricity prices, it is essentially asking NEA to provide them a guarantee similar to IPPs asking for a USD PPA.  If that is the case, then it should also be willing to pay for the cost associated with it. I do not believe that it would be very difficult to structure a price mechanism whereby price increases are kept to minimum or zero for low-end user and prices rise rapidly as usage increase.  Essentially if you want to use more electricity then you will have to pay more with the marginal cost of electricity per unit rising much more rapidly with each threshold. 

Finally, I would like to note that while I have stated that USD equity investment should not be considered when determining the ratio on which to apply the USD rate for electricity, this should not be construed that I am against it.  It can be incorporated if it is in the interest of the country to see the project go through.  I am pretty sure that if we all put our heads together that we can easily come to a solution that can solve our electricity crisis starting with how a PPA is structured.  There is no need to be for or against it.  A market mechanism can be constructed to create a win-win situation for all.