What Drives FPI Inflows into the Indian Debt Market?

Lets begin this post by being a little pedantic, shall we? But before that, inline with tradition (we love nerdy humour on nerdverve), let’s poke some fun at pedants like us.

I present to you, exhibit A: “Gary the Grammar Cactus Will Never Find Love”.

This was forwarded to me by a friend (you know who you are) and made me laugh for days! Now that we’ve established that I will never find love, lets get down to business.

FPI vs. FII: Lets get pedantic

FII or Foreign Institutional Investor refers to an institution established or incorporated outside India, which proposes to make investment in securities in India. While the term “FII” was used commonly before, it has been replaced by FPI or Foreign Portfolio Investor now. Starting June 2014, a new investor class called FPI was created by SEBI by merging the 3 existing foreign investor classes - FIIs, Sub Accounts (of FIIs) and Qualified Foreign Investors (QFIs). Note: “Sub Accounts” include foreign corporates, foreign individuals, those institutions established or incorporated outside India and those funds/portfolios established outside India (incorporated or not), on whose behalf investments are proposed to be made in India by a Foreign Institutional Investor (FII). QFIs include foreign individuals, groups or associations resident in certain countries.

In common parlance (even in business publications actually), FII and FPI still tend to be used interchangeably, even though the correct term to use is FPI. Also note, the data referenced in Indian business publication is always FPI data (provided by CDSL and NSDL) even when referred to as FII data.

Remember, foreign portfolio investment by any single investor cannot exceed 10% of the equity of an Indian company, beyond which it is treated as FDI (Foreign Direct investment).  Total limit for FPI in an Indian company is 24% (of total equity) – this limit can be raised up to the FPI cap for the sector.

While I’ll talk about FDI in later posts, I’m going to quickly mention the key differences between FPI and FDI before we move forward. FPI refers to smaller stakes in Indian companies. FPI investors are usually passive, are not involved in management and strategy decisions of the company, and have a relatively shorter investment horizon. FDI investors on the other hand usually have controlling stakes in Indian companies, are involved in management decisions and tend to be long-term investors.

What drives FPI inflows into the Indian debt market?

The factors that drive FPI flows into the Indian debt markets are essentially the same that drive FPI flows into any debt market. I decided to focus on the motherland in this post simply to make this discussion topical and allow you to say at least a few smart things when you see your friends next.

I’m going to refer to the Jan-Apr 2015 period to underline these drivers.  (Our next post will deal with factors that are currently driving FPI flows out of India – what a difference a few months makes!)

Jan – Apr 2015: Huge FPI Inflows into Indian debt

The Jan-Apr period of this year saw huge FPI net inflows into the debt market (Rs 20,000+ Crores in Jan, Rs 13,000+ Crores in Feb, Rs 8,600+ Crores in Mar). A number of favorable factors combined to create this welcome state of affairs.

Note: Net FPI inflows for July were just USD 4 million, and for Sep till date, the figure is USD 227 million. These values don't show up in the graph above because of the scale.

·     Higher yields vs. developed markets. Bond yields (=coupon received/price of bond) in India are much higher vs. those in many developed countries. In the Jan-Apr period, the 10-yr US Treasury yield averaged 1.95%, while the 10-yr Indian government bond yield averaged 7.76%. The interest rate differential was 580 bps! (See charts below). Of course this higher yield had to be weighed against the higher risk that Indian government bonds embody vs. US treasury bonds. Per S&P, India’s sovereign credit rating is BBB- against the US’s AA+. The higher returns however were quite attractive early in the year despite the higher risk, due to a number of other favourable factors mentioned below.

·    Falling interest rate environment = mark-to-market gains. There are 2 ways investors make money on bonds. First is through the regular coupons payments they receive. This is called “carry” (i.e. yield of bond). The second is through mark-to-market (MTM) gains or capital gains. This happens when interest rates/bond yields fall, which causes bond prices to rise since bond price is inversely relate to yield (we’ll do a separate post on this sometime I promise). A rise in bond price means capital gains or MTM gains for the investor.

In the Jan – Apr 2015 period, the RBI cut the repo rate twice (Jan and Mar) by 25 bps each (to 7.5%), with more cuts expected by the market in the coming months. As a result, bond investors poured money into the debt market hoping to profit from capital gains.

·      Relatively strong, stable Rupee.  The rupee was amongst the best performing currencies viz-a-viz the dollar in the Jan-Apr 2015 period. A stable government with a reform agenda, improved macro expectations, a smaller current account deficit, falling inflation trend and expectations of rate cuts from the RBI, all led to strong FPI inflows (debt + equity) into the country. Given this backdrop, the RBI bought a massive amount of dollars to prevent the rupee from appreciating.

Let me explain how this works. When there are strong FPI inflows into the country, the supply of dollars increases vs. the available rupee supply, causing the rupee to appreciate vs. the dollar. This is bad for the Indian export industry since a stronger rupee tends to make exports more expensive for foreign buyers. In order to avoid this situation and prevent exchange rate volatility, the RBI usually buys dollars. This curbs excessive rupee appreciation and keeps the USD-INR exchange rate stable/within a range.

However, this tends to be inflationary. Why you ask? The RBI prints rupees in order to buy dollars thereby increasing the available rupee supply in the system. More money chasing the same amount of goods leads to inflation. To avoid this unintended consequence, the RBI does what is called “Sterilization”. Sterilization refers to the issuance of government bonds by the RBI to mop up the excess rupees released into the system when it buys dollars/ intervenes in the forex market.

Anyways, coming back to point, given huge FPI inflows into the Indian debt and equity markets in the Jan-Apr 2015 period, the RBI bought a massive amount of dollars in the same period, to prevent excessive appreciation of the rupee vs. the dollar. This kept the currency stable and was reassuring for foreign investors, who continued to pump in more money.

           

·      Steady Fed Funds rate and other (India-specific) factors.  While we’ve talked about the key drivers that helped attract FPI inflows into Indian debt earlier this year, there are many other factors that FPI investors consider while allocating funds. One of the most important is the Fed funds rate trajectory. In the Jan-Apr period, the Fed funds rate was steady with a hike not expected before June. When the Fed raises rates, interest rates in the US rise and the dollar appreciates, causing the interest rate arbitrage opportunity between India and US to narrow, which leads to FPI outflow from Indian debt.

Other India - specific factors such as strong macro fundamentals (vs. other emerging markets) and GDP growth, falling inflation trend, stable government and reform agenda etc. also played an important role in driving FPI investment into Indian debt during the Jan-Apr 2015 period.  

We’re going to end here, and talk about factors engendering FPI outflows (the situation today) in our next post.  

Term Repos and the Development of the Money Market in India

After explaining how overnight Repos work in the last couple posts, in this post, I'm going to talk about RBI's move towards "Term" Repos in their endeavour to develop and deepen India's money market. 

Unlimited access to cheap funds from RBI till Mid-2013

Under the RBI’s Liquidity Adjustment Facility (LAF), banks had unlimited access to cheap funds through overnight repos till mid-2013. As a result, instead of raising short-term money through deposits, or borrowing in the inter-bank call market or through other money market instruments, banks would end up borrowing most of what they needed from the LAF overnight window at the fixed Repo rate set (7.25% currently) by the RBI. Why bother with anything else when you could get the cheapest funds here?

Why was this a problem?     

• Thwarted the development of money market yield curve: Being able to borrow an unlimited amount of overnight funds through the LAF window meant that banks could borrow overnight, and meet longer-term liquidity and funding needs by rolling over the repo collateral. In effect, they were borrowing longer-term funds at the overnight rate. This is a problem because when banks don’t need to borrow for longer terms (7 days, 14 days, 28 days and so on), they don’t learn how to price risk for such maturities which means they don’t understand how much interest to pay for such maturities. This thwarts the development of a money market yield curve (“Yield curve” is simply a graph where the yield of fixed-interest securities is plotted against their maturity) which is never a good thing. A deep, mature money market requires there to be an organized, developed and liquid market for instruments of various maturities where participants can confidently transact at rates/yields that move within a tight range.

• Banks using LAF to fund credit growth: Banks were using cheap overnight LAF funds for driving credit growth, rather than focusing on accelerating deposit growth in order to drive their credit portfolio. It was a no-brainer for them since the Repo rate was lower than the interest they’d have to pay to depositors. That said, this is unhealthy for the banking system on a fundamental level. The rates at which banks lend to customers should be determined by the rates they have to pay to depositors (along with other factors) for the same tenors. Having unlimited access to cheap funds can lead to excessive credit growth without the necessary due-diligence.

• No incentive for banks to manage liquidity needs better: Given that they could access whatever funds they needed through the LAF window, there was no real incentive for banks to better forecast and manage their short-term liquidity needs.

Starting July 2013, RBI began to restrict access to overnight funds; Introduced Term Repos in Oct 2013

·    July 15, 2013 - the RBI restricted overnight funds borrowed under the LAF to 1% of NDTL (Net demand and time liabilities) of the banking system (1% of NDTL was ~Rs 75,000 crores).

·     July 24, 2013 - the RBI cut overnight LAF borrowing limit again to 0.5% of NTDL (0.5% of NDTL was ~ Rs 38,000 crores).   

·    Oct 7, 2013 – the RBI introduced 7-day and 14-day “term” Repos, the borrowing limit under which was set at 0.25% of NDTL, to provided additional liquidity (in addition to overnight repo borrowing limit of 0.5% of NDTL). Variable rate auctions would be conducted for these term repos where banks would put in their bids expressing the term repo rate they were prepared to pay. The 14-day repo auction would be conducted every reporting Friday, while the 7-day repo auction would be conducted every non-reporting Friday.

·    Oct 29, 2013 - the RBI increased liquidity provided under term Repos to 0.5% of NTDL (~ Rs 40,000 crore) from 0.25% of NDTL.

·   Apr 1, 2014 – the RBI increased the liquidity provided under term repos from 0.5% of NDTL to 0.75% of NDTL (which was ~ Rs 60,000 crore). Liquidity provided under overnight repos was reduced from 0.5% of NDTL to 0.25% of NDTL (which was ~Rs 20,000 crore).  

What was the impact of these changes?

Lets recap - so, from July 2013 to April 2014, RBI had cut the unlimited overnight repo borrowing under the LAF window down to 0.25% of NDTL (i.e. from the Rs 100,000+ crore overnight borrowing in early 2013, the overnight borrowing limit was down to  ~Rs 20,000 crore in Apr 2014).  

Liquidity provided under term repos was increased from 0.25% of NTDL in Oct 2013  (~Rs 20,000 crore) to 0.75% of NDTL (~Rs 60,000 crore) in Apr 2014.  

So, the total liquidity support available through overnight and term repos was 1% of NDTL (~Rs 80,000 crore).

What was the impact of these measures?

• Firstly, bank began to start borrowing from the RBI at higher “term” repo rates. Maybe 20-50 bps higher vs. the fixed overnight repo rate they were borrowing at earlier. 
• With overnight LAF repo borrowing cut drastically, banks had trouble managing liquidity. As a result, call rates became volatile and fluctuated for some time in a rather broad range of ~100 bps. 
• In the 14-day Term repo auctions, banks would borrow everything they could, because they weren’t really sure how much they would need. If they had extra funds, they’d then lend them in the call market. This however, also led to call rate volatility. Note: when banks borrow from the RBI at the “term” repo rate, and then lend these funds in the call market at the call rate (RBI wants this to be close to the overnight repo rate) or do a reverse repo with the RBI at the reverse repo rate (repo rate – 100bps), they tend to lose money! This is usually not the way to go, unless the market is experiencing volatility. 

Basically, banks had trouble managing their liquidity needs, which manifested as volatility in call market rates.

Starting Aug 2014, RBI started conducting more frequent Term Repo auctions for better liquidity management and stable Call rates

On Aug 22, 2014, The RBI announced that it would now conduct variable rate, 14-day repo auctions 4 times in the fortnightly reporting cycle (on Tuesdays and Fridays). The amount auctioned under each 14-day repo auction would be equal to ¼^th of 0.75% of NDTL.  So, while the amount of liquidity offered under term repos every fortnight remained the same (0.75% of NDTL), it was now split into 4 tranches. This increase in frequency was to allow banks to better manage their liquidity needs, so that call market rates remained to close the Repo rate (as the RBI desires).  In fact, the RBI governor has often said that he wants the overnight call rate to “hug” the repo rate.

Why does the RBI want the overnight Call rate to “hug” to the Repo rate?

The RBI should ideally be the lender of “last resort” to the banking system. Banks should raise money from deposits, money market instruments and borrow from each other (call money), before they turn to the RBI. The call money market is where banks lend/borrow from each other overnight (no collateral security required).

However, in India, the RBI has been (and still is to large extent) the lender to whom banks tend to go first, because they offer Short term funds to banks at the cheapest rate. The RBI wants the call money market (overnight) rate to “hug” the repo rate, so that banks borrow more from each other in the call market rather than depending on the RBI for most of their needs. If the call rate remains stable and close to the Repo rate, the call market will automatically develop and deepen. If there is constant volatility in call rates, this will be hard to achieve.

How is the liquidity situation now?

The banking system has been flush with liquidity over the last 3 months due to higher government spending, lower credit growth and  increased dollar buying by the RBI. As a result, the RBI has been trying to mop up the excess liquidity in order to keep the call rate and other market rates in check and curb inflationary pressures.

It has been conducting overnight reverse repo and term reverse repo auctions (5-14 days) where banks place funds with the RBI and earn interest at the reverse repo rate (reverse repo rate = repo rate – 100bps; currently = 6.25%) and term reverse repo rate (decided at variable term repo auctions) respectively. The RBI has also conducted Open Market Operations (OMOs) where it sells government bonds to banks, thereby sucking in some liquidity.

If the RBI didn’t actively manage excess liquidity in this way, call rate would have fallen sharply and inflationary pressures would intensify. 

Are banks better at managing liquidity post these moves by the RBI?

While banks still have a long way to go as far as forecasting short-term liquidity needs and optimal liquidity management is concerned, these moves by the RBI are forcing them to move in the right direction.  The deepening and maturing of the money market in India and the development of a robust money market yield curve will take time. That said, the RBI is on the right track.  

How Repos Work - Introducing The “Haircut”

Now that we know how a Repo transaction actually works (read my previous post "How Repos Work - Some Uncomplicated Excel Sheet Indulgence”), we’re ready to talk about “haircuts”.

While these haircuts are certainly NOT the kind your stylist gives you, I’d imagine they have the same re-assuring impact on repo buyers as my pixie cut has on me (I look like a homeless teenage boy without it).

Think I’m embellishing?

Allow me to present Exhibit A.

Homeless teenage scamp or kooky mad professor?? Hmmm......I can never quite decide.

But I digress.

The hero of this post is the “haircut” that repo buyers (lenders) impose on the market price of collateral securities in order to protect themselves in case of a default by the seller (borrower). Lets assume that the buyer imposes a haircut of 2% on collateral securities whose market value is Rs 50 crores. This means that the buyer will pay the seller 98% of the market value i.e. Rs 49 crores in cash for the securities. At the end of the Repo term, the seller will return the Rs 49 crore to the buyer along with the repo interest on this amount, while the buyer will return the securities worth Rs 50 crore in market value (assuming their price hasn’t changed) back to the seller.  

In a regular repo transaction (sans haircut), the seller would have paid the buyer the full market value of Rs 50 crore in the first leg of the repo transaction. The reason many repo buyers insist on a haircut is to protect themselves from any loss they may incur when they try and sell the collateral securities in case of a default by the seller.

In the example above, lets assume that the seller defaults while the price of the collateral securities falls to Rs 49.2 crores during the term of the repo. In such a situation, due to the cushion of the haircut, the buyer can still sell the securities and make a maximum return of Rs 20 lakh (Rs 49.2 – 49.0 crore) on the deal – the actual return will be less due to transaction costs. If the price of the securities falls to Rs 49 crores, the buyer may at least prevent a loss. 

Besides the price volatility of collateral securities, there are other risks that a “haircut” is designed to protect the seller against, such as:

1)   Liquidity risk: difficulty in liquidating securities without loss in value (finding buyers of sufficient quantity, market impact of big sales etc.)

2)  Legal risk: delay in selling of securities due to litigation challenging the right of the buyer to sell collateral securities post default.

Now that we understand Repo haircut basics, lets do what we love best - a little real-world math! * nerd grin *

The excel sheet below depicts a regular Repo Transaction (without a haircut).  It’s quite self-explanatory.

The sheet below shows the same Repo transaction with a 2% haircut.

The difference between both worksheets is that for the same face value (Rs 50 crore) and market value  (Rs 56.17 crore) of collateral bonds, sans haircut the buyer pays the full market value (Rs 56.17 crore) to the seller, while with a 2% haircut, he pays 98% of the market value of the bonds (Rs 55.04 crore) to the seller. In both cases, the buyer earns 6% in annualized repo interest.

Lets now see how the haircut protects buyers in case of a default where the collateral securities lose market value. Let’s assume the clean price (excludes accrued interest) of the collateral bonds falls from Rs 109 to Rs 108. The worksheet below shows that without the haircut, the buyer will loss money (~Rs 34 lakhs) and make a negative return of 15.8% annualized. With the 2% haircut however, the buyer will make a big return (36.4% annualized) despite the fall in price. Note: this is just an example – potential profits / losses are seldom this big.

The price of the collateral bonds will have to fall to below Rs 106.44 (from Rs 109.0) for the buyer to make a loss with a 2% haircut. See worksheet below. 

Despite my love-affair with excel sheets, I'm beginning to feel like we'd had enough of them for one post. Also, it's time for me to go to the salon. Ciao.