Pakistan's Debt Dynamics: Less External Debt, Please!

Every once in a while, an ‘expert’ will expound their belief in various news media that Pakistan’s debt, especially domestic debt, should be restructured. What’s the key evidence or theory backing their arguments? Nominal figures stating the government could save an X billion amount of dollars in debt servicing.

These flimsy arguments are comical, to say the least. Another frequently cited statistic is that each 100-bps cut in interest rates saves the Pakistani government an estimated Rs 250bn. The interpretation of these figures totally misses the mark and primarily suffer from money illusion. Debt and its servicing costs cannot and should not be seen in nominal terms, alone.

Firstly, debt only has to be restructured when the borrower is unable to fulfil its financial obligations. When a borrower faces difficulty in making debt payments, creditors give concessions entailing a change in the payment terms (reduction in interest rate or extension of maturity) or a write-down (partial waive off) on the principal. This is done to avoid a costly bankruptcy case in the event of a default.

When it comes to domestic debt, the government is not close to this scenario at all as the SBP provides adequate liquidity to meet the government’s borrowing needs. Moreover, commercial banks are happy to continue funding government borrowing.

The need for domestic debt restructuring only arises if banks (or other financial institutions) decide not to fund government borrowing which could only happen due to a crisis of confidence. Historically, there has never been domestic debt restructuring which is why that crisis of confidence has not arrived.

However, hypothetically, if the government were to restructure domestic debt today, a crisis of confidence could arise in a future high debt-to-GDP ratio scenario. Which is why talks of domestic debt restructuring must not be taken lightly.

But what about the factor that really matters? Debt dynamics of domestic and external debt work differently for several reasons.

The above equation shows the domestic debt dynamics. The change in domestic debt-to-GDP ratio depends on a few key variables – real interest rate, growth rate, and primary balance. If r > g, debt-to-GDP escalates. Meanwhile, if the primary balance is negative (i.e. a primary deficit), then the debt-to-GDP worsens too.

Historically, real interest rates have averaged out around 0.1%-2.7% (depending on the choice of time period), while long-run growth rate of the economy has been 4.7%. So, in Pakistan, g > r by far which is positive for domestic debt dynamics. Implicit in this is the fact that the government has deflated its domestic debt over time – r has been low due to high inflation and central bank non-independence. Meanwhile, primary balance has been -0.5% of GDP since 1992 – which is not a good thing (IMF).

However, when you run these figures together in the equation, the total impact on domestic debt-to-GDP ratio has been negative – meaning domestic debt is sustainable or falling (in real terms).

Among this, there is an assumption that government revenues grow at the rate of nominal growth rate of the economy. Even that assumption mostly holds as nominal growth rate has been 13% while tax revenues have grown 12% over the past decade, for instance.

When it comes to external debt, the situation becomes murkier. To pay off external debt, the government does not only require tax revenue but adequate foreign reserves. Moreover, exchange rate shocks can escalate external debt payments (in local currency terms).

In the above equation which shows external debt dynamics, pcb stands for primary current account balance, while △res stands for change in reserves. In Pakistan’s case, interest rate on foreign debt r* has been 3.6% historically – much higher than that on domestic debt. Furthermore, the particularly troubling part is that this is only brought down due to concessional multi or bilateral lending.

The interest rate on commercial borrowing and bonds are usually 5% or above even in supposedly good times. Moving on, primary current account balance has been -1.9% of GDP since 1980. Even if you take out the effect of change in reserves (which is likely negative), the impact on external debt-to-GDP turns out to be positive – meaning external debt is rising in real terms or unsustainable.

This is some simple arithmetic which tell the true story of Pakistan’s debt dynamics. This is why when the Pakistani government restructured its debt in the 1970s and early 2000s, only external debt was restructured and not domestic debt. The math and some history speak for itself.

Cement Cos: Who is Best Poised to Benefit from Rate Cuts?

Cyclicals on the PSX have been on a serious bull run this year. Certain sectors such as the auto and fertilizer sectors have been on fire. Why? Well, expectations of an interest rate cut are leading to the belief that demand in the economy will pick up. Cyclicals would be a natural beneficiary.

The cement sector is another cyclical sector which has been at the forefront of this rally. With capacity utilization expected to be at 57% for FY24 and more capacity coming online over the next year, cement companies could see significant growth if demand does rebound.

But which cement companies would be best suited to exploit this opportunity? Or which ones could outperform the pack?

The important thing to note about cement is that the product is quite identical in quality, regardless of the producer. Moreover, the price of cement varies only based on the location of each producer (transaction cost). So, the differentiation in cement producers comes down to the production cost of each producer and their capital structure (debt-equity mix).

As a consequence, the key determinant of a cement company’s market value should be their production capacity.

If the market is valuing the production capacity of one producer more than another, it may be due to certain idiosyncratic characteristics. For instance, Bestway Cement has better governance practices than its competitors, so it naturally trades at a premium. But on average, companies with a lower market cap per ton of production capacity should be seen as cheaper on a fundamental basis.

Similarly, those with a lower PEG than the average should be seen as attractive as an investor is able to purchase growth at a lower per unit price. For the same reasons, a lower-than-average price-to-book value and EV per EBITDA would be attractive. However, a lower EV per EBITDA is special for another reason too.

Enterprise value is simply the total value of the firm’s debt and equity. Once interest rates fall, value would be transferred from debtholders to equity owners as finance costs for a company would fall. Or in other words, income flow to debtholders would fall.

Thus, a lower EV per EBITDA implies that the market is pricing each rupee of enterprise value for each rupee of earnings at a low rate. When interest rates fall, that enterprise value would become more valuable for shareholders.

On the flip side, a better-than-average return on equity and assets would be preferred as it implies more efficient capital allocation.

These metrics have been calculated here for the cement sector. Based on these metrics, Maple Leaf Cement looks the cheapest. Other seemingly cheap companies appear to be Fauji, Attock, Cherat and Kohat Cement.

Lucky Cement also seems to look cheap, however one cannot treat it as a true cement company due to its diversified (conglomerate) nature.

Based on these calculations, personally I prefer Maple Leaf, Attock and Fauji Cement among these fundamentally cheap companies, as they have a lower-than-average market cap per ton of production capacity.

An alternative viewpoint could also be to consider debt-to-equity ratios and pick those companies with the highest debt-to-equity ratios. Why? Well, when interest rates fall, finance costs (interest expenses) for such companies would plummet allowing them to enhance their earnings.

However, I personally do not subscribe to this view as it seems to take on unnecessary risk. One must note that the cement companies with the highest debt-to-equity ratios are those who are loss-making. What does this say about the management of these firms? Plus, why take on needless interest rate risk?

Furthermore, EV per EBITDA as a metric already captures the element of falling interest rates buttressing the financial viability of a firm.

So, based on this, when can one expect some of these cement shares to catapult themselves to new heights? Well, that’s the catch; one needs to be patient. It was only last week that Gharibal Cement (GWLC) was looking most cheap on these metrics.

Yet, within a week the share price of the company has gone up 37%. Now suddenly, the company looks realistically priced on a relative basis. That’s the thing about markets. They ignore value for some time. And when they do notice value on the table, they do so with aggression.  

Why OMOs Don’t Always Cause Inflation

Many commentators have been touting the monetarist belief that increased money supply through SBP’s Open Market Operations (OMOs) will continue to stoke inflation. It’s a belief I’ve held too but it is not true as circumstances have changed.

Money supply has been rising steadily over the years with an uptick since 2021. Monetarists would tell you this has been the true cause of inflation in Pakistan while quoting Milton Friedman: “Inflation is always and everywhere a monetary phenomenon”. 

(Source: State Bank of Pakistan)

This belief stems from the fundamental monetarist equation MV=PY (where M stands for Money Supply, V for Velocity of Money, P for Price Level and Y for Real GDP). A monetarist will tell you that if you hold V and Y constant, the only cause of inflation would be money supply growth. Thus, curtailing money supply would curtail inflation. Meaning, money supply growth and inflation (price level growth) follow a one-to-one relationship.

Generally, this has been true as money supply (M2) growth has largely tracked inflation in Pakistan. However, there has been a differential between the two much of the time. This is because the monetarist assumption of Y (and V) being constant does not always hold. 

(Source: IMF) 

For instance, most of the 2000s and early 2010s were a period of positive supply shocks marked by low commodity prices, thus helping generate lower inflation than as predicted by M2 growth. But this can work in the opposite direction too as the severe import restrictions and commodity price shocks in 2008-2009 and 2022-2023 led to inflation spiking beyond M2 growth. 

Nonetheless, a general trend between money supply growth and inflation can be inferred. So even if we take the monetarist belief as the rule and these slight divergences as the exception, how does SBP create money? 

For one, SBP can circulate more notes and coins to increase money supply. This is generally done to meet demand for money. A standard example is that during Ramadan/Eid, money demand spikes so the SBP caters to this demand in order to keep interest rates the same. And also, to not curtail real activity in the economy. After all, money is there to facilitate real transactions.

But the more sophisticated approach to creating money involves open market operations. Outright open market operations generally entail the SBP buying or selling bonds to vary money supply. Generally-speaking, when SBP buys bonds, the money supply goes up as the public has more cash while SBP selling bonds does the reverse. Note, broad money supply is defined generally defined as currency in circulation plus short-term deposits (current and savings account for instance). 

 

The transaction involves SBP buying bonds from financial institutions and crediting the accounts of those institutions. Now, ideally banks would use that credit to lend it out to borrowers as they do under normal circumstances. Those borrowers would keep their loans in the bank (as deposits), thus creating money and having an impact on the real economy. Note, in this process, it is the banks that are actually creating money through their choice to lend or not although SBP began the initial process of credit. 

However, when economic conditions worsen, banks generally tend to lend to the government or simply hold excess reserves. So greater credit (monetary base) does not translate to greater money supply. Hence, understanding how government borrowing is facilitated and how that money is spent can help unravel inflationary trends. 

When the government is short of cash, SBP performs what is known as reverse repos (another kind of OMO). In a reverse repo, SBP will buy T-bills or PIBs from banks with the agreement to purchase it at a higher price at a future date. The difference between the future purchase price and current sale price is determined by the interest rate. Meanwhile, the future purchase date depends on the tenor set by the SBP which usually ranges from 7 to 28 days. 

Now, when the government needs cash, SBP will perform these reverse repos so that banks have more cash. So, when the SBP holds auctions for government securities (to raise cash for the government), the banks will fund it with the cash received from the reverse repo. This is the great monetary settlement between the SBP, the government and the commercial banks.

But hang on a second, how does this cause inflation or affect the real economy? Well, when the government receives that money from the auctions, it chooses to spend it on goods and services. So, let’s say the government buys wheat from flour mills for a new subsidy program. The flour mills receive that cash and choose to spend a portion of it (or hold a portion as deposits). That spending leads to an income for someone else and this cycle continues ad infinitum. 

This is precisely how SBP’s reverse repos lead to inflation. But implicit in this example is that the government is spending that cash on goods which affects the real economy. As in, it is running a primary deficit which it generally has for most recent years.

However, this financial year, the government is actually running a primary surplus. Meaning, it is not really spending that reserve repo cash on goods. So where is that cash going? Well, it’s not really going anywhere; it’s just being circled around.

Essentially, that cash is being spent on debt servicing. So, the SBP is still handing cash to the banks which are buying government securities at auctions and giving the government that cash but only to retire or service old debt. So really, the money is not being spent anywhere – not making its way into the real economy or increasing the number of deposits. 

In this scenario, OMOs really do not cause inflation. An interesting example or parallel in monetary history to this would be the German MEFO bills of the Nazi-era. 

During the Nazi-era, the German government needed a lot of cash to reinflate the economy. Naturally that would cause inflation, so Hjalmar Schadt (the German Central Bank President) introduced an ingenious security called MEFO bills.

MEFO bills were promissory notes that served as bills for payment by the German government to manufacturers. They were convertible to the German currency but a higher interest rate on them was offered to entice manufacturers to hold those notes. Meanwhile, the maturity on these notes would continually to be extended so as to avoid conversion of these notes to the currency – to avoid increasing money supply.

So, what’s the lesson here? Greater credit does not always mean greater money supply or inflation. Consequently, OMOs (credit creation) do not always cause inflation. At least, they are not causing inflation in the present circumstances. 

KSE100: The Market Cycle and the Macros

The Pakistani stock market is known to follow a well-established market cycle trend. It booms for a few years and then retreats or stagnates for a few more. Rinse and repeat ad infinitum. As covered in a previous blog, macroeconomic fundamentals are the key trigger for this behaviour; some of which include oil prices and the PKR/USD rate, or rather the stability of the rate. 

Over the long run, the stock market should appreciate to provide a positive real return in dollar terms (theoretically, at least). The graph below plots out the KSE100’s dollar market cap over time along with the 5-year moving average. Since the 5-year moving average is a backward-looking lag variable, it tends to be sticky and trails the market cap. 

Based on the graph, one can quickly notice two periods of 1997-2002 and 2009-2012 where the 5-year moving average faced a slump. Shortly after, the dollar market cap of the KSE100 index picks up. Note, the 5-year moving average lags the improvement in market cap by roughly a year. 

Now from 2018-2023, the 5-year moving average faced a similar slump. But note, 2023 saw the KSE100 appreciate 24% in dollar terms. Hence, the five-year moving average is likely to pick up from next year. 

Once the five-year moving average picks up, it remains on the uptrend for about six to seven years. Once again, the lagged nature of the variable means it lags the start and end of the rise of the KSE100 dollar market cap by a year. What does this mean? The market is looking very tasty for the next three to four years both in PKR and USD terms.

Interestingly, in each uptrend cycle, the previous dollar market cap high is broken. With the current market capitalization of KSE100 standing at $34 bn, breaking 2017's high of $91 bn would imply a 167% return (in $ terms). Note, the all-time high KSE100 dollar market cap is actually $99 bn.

However, this might be improbable given the steepness and depth of the preceding market decline since 2018 as well as the massive overvaluation of the rupee in 2017. Nonetheless, the stock market should yield solid returns over the next few years.

Moving on, the graph above uses a metric known as the 'Buffett Indicator'. Named after Warren Buffett who used to tout its efficacy, the metric utilizes the fact that the stock market must reflect changes (growth) in the GDP, thus the indicator must be constant over time. 

Based on this, the KSE100's Market Cap as a % of GDP has been far below its long-run average over the past few years once again highlighting the cheapness of the market at present terms. Once the market enters the boom phase of the cycle, the Buffett Indicator tends to remain above the average for roughly four years. 

The two graphs above display the relationship between the KSE100 index and brent crude price as well as the PKR/USD parity. 

With the PKR/USD parity, we observe that when the rupee is stable, the KSE100 tends to perform well. This can be rationalized by the fact that when the Pakistani rupee is stable, investors are confident that the value of their investment will be sheltered. This is of great importance to foreign investors which is why foreign portfolio investment rises during periods of PKR stability. But local investors also shift their capital to rupee-denominated assets like Pakistani stocks during such periods as dollar-denominated assets become less attractive in rupee terms.

At the same time, it is seen that the KSE100 tends to fully reflect and adjust for PKR depreciation. In fact, the graph shows that KSE100 far outperforms holding the USD. Since the inception of the KSE100 index, the index has provided a cumulative return of 2666% versus the dollar’s cumulative return of 813%. 

As for oil prices, the KSE100 index tends to perform well when brent crude prices are falling or stable. The 2010s were notable as brent crude prices crashed but the KSE100 index soared with that result. The transmission channel here is that lower brent crude prices improves Pakistan’s current account which lowers pressure on its foreign reserves, thus helping the country avoid (or at least delay) a balance of payment crisis. This in turn relieves speculative pressure on the PKR/USD rate. 

However, KSE100 can perform well even in the face of rising brent crude prices as the 2000s showed. However, the prolonged period of rising crude oil prices eventually did push Pakistan into a balance of payment crisis in 2008-09. So, in general, falling brent crude prices are favourable. 

However, oil price changes are a major contributor to price level changes so over the long run, the KSE100 index tends to show a positive relationship with brent crude prices. This is because the index will and does adjust to reflect price level changes over the long run. 

What does all this mean? The Pakistani stock market is a great buy at the moment despite the spectacular rise it saw last year. The uptrend phase of the market cycle has just begun, and Pakistan’s macros look good. The key question would be: when should one exit? The good news is you probably have a few years to contemplate that question. 

Allocating across Asset Classes: Gold, USD or Stocks?

Over the past few years, a large segment of the Pakistani population has espoused the belief that the dollar is the best asset class in Pakistan. Traditional economic theory would differ from this belief as prices and exchange rates have a one-to-one relationship in the long run, thus holding the vehicle currency (dollar) over the long run provides no real return. But let’s take a deep dive into how gold, USD and stocks perform over time.

Firstly, Pakistani fixed income investments are being omitted from the analysis as they’ve barely provided a positive real return over the long run; inflation has averaged around 7.9% over the long run while interest rates have averaged around 8%. Moreover, real estate has been excluded due to lack of aggregative data on property prices. Zameen.com does have a property price index but it is limited in its time frame as well as geographical aggregation.

Historically, gold has been expounded as an inflation-neutralizing asset class. However, that statement applies when the comparison is made for developed nations. In Pakistan, it is generally an inflation-beating asset class. This is a result of gold being a dollar-denominated asset so not only does its return reflect depreciation/devaluation of PKR but also increase in the price of gold over time. The twofold source of return provides a nice cushion for a gold investor.

As for cross-comparison of asset classes versus inflation, it is observed that the USD has not beaten inflation over a long-time frame. Meanwhile, both the KSE100 index and gold have significantly outpaced inflation. Till the end of 2023, gold edged out the KSE100 index but the trajectory of both stocks and gold in Pakistan is fascinating, to say the least.

Interestingly, there have been time periods where the KSE100 index has outperformed gold but also a huge stretch of time where it has lagged behind gold. Much of this is down to the market cycles observed in stocks and real estate in Pakistan.

Based on these observations, an astute investor would not allocate any of their capital towards the USD as gold is a much better inflation hedge which already incorporates the return from holding the USD. But to take it one step further, allocation between gold and stocks should be tinkered with from time to time.

When the market cycle enters the growth phase, one should shift their allocation towards stocks as stocks tend to outperform gold during this phase. Meanwhile, when the market cycle returns to the stagnation phase, one must lean more towards gold in their allocation.

It is important to note that correlation between gold prices and the KSE100 is 0.68 suggesting a moderate positive relationship. Thus, reflecting the fact that both asset classes adjust to nominal inflation effectively over the long run.

My personal gripe with real estate in Pakistan is the lack of liquidity and the large amount of investment required to invest in real estate – which leads to overallocation of capital to real estate. Nonetheless, selective pockets of real estate in the major cities of Pakistan have been excellent investment opportunities. If one possesses information not available to the public, plenty of money can be made in real estate. But that means high barriers to entry so not something a regular investor can count on. Thus, the emphasis on gold and stocks remains.

Regardless of the viability of real estate as an asset class, this much is clear: holding the USD makes for a very poor investment. One should stick to a mix of gold and stocks. Gold can be a great inflation/depreciation hedge, while stocks can act as a good source of steady income (dividends) and inflation hedge over the long run.