Tag Archives: Turkey

Turkey’s Attempt to Mop up Mattress Gold

This blog post is a guest post on BullionStar's Blog by the renowned blogger JP Koning who will be writing about monetary economics, central banking and gold. BullionStar does not endorse or oppose the opinions presented but encourage a healthy debate.

The Turkish government, in particular its central bank, has a relatively unique relationship to gold. Few governments directly involve themselves in the gold market as much as Turkey's. This post provides a short review of this relationship.

Gold is popular in Turkey. According to a 2015 report by the World Gold Council, Turkey generally accounts for around 6% of global gold consumer demand. It is further estimated that there are around 3,500 tons of under-the-mattress gold in Turkey, which makes Turkey one of the largest gold sinks in the world. The existence of this "mattress gold" is believed to inhibit the growth of bank intermediation. By encouraging Turks to bring their gold to banks to be deposited, Turkish authorities believe that overall engagement with banks will increase, presumably leading to faster economic growth. Notably, India has also been very aggressive in trying to bring "idle" gold into the banking system, having launched a gold monetization scheme in 2015.

Required reserves in the form of gold

The nation's central bank, the Central Bank of the Republic of Turkey (CBRT), has become particularly active in the campaign to draw gold into the banking system. It does so by allowing central banks to keep some portion of their required reserves in the form of gold. Let's unpack the idea of reserve requirements a bit more. Central banks often (though not always) require deposit-taking commercial banks to permanently freeze a fixed amount of reserves at the central bank. (Canada for instance has no reserve requirements). A central bank that imposes reserve requirements typically remunerates reserves at a below-market rate, sometimes 0%. Required reserves therefore function as a tax on the banking system. If banks weren't required to keep funds in their central bank accounts, they would be investing the funds elsewhere at a higher rate.

Turkish banks were cut some slack in 2011 when the CBRT implemented a special program called the Reserve Option Mechanism, or ROM. Instead of requiring that banks meet their entire reserve requirement by holding poorly-compensated Turkish lira funds at the CBRT, the Reserve Option Mechanism allowed them to satisfy up to 30% of their requirements by depositing gold at the central bank. This offer was quickly taken up by Turkish banks, who began to deposit gold up to the maximum allowed amount. The chart below  illustrates the popularity of the ROM. The 30% tranche, which was introduced in stages starting with a 10% tranche in 2011 and subsequently boosted to 20% and then 25% in 2012, has almost always been used to its full capacity.

Utilization of the CBRT's Gold Reserve Option Mechanism (Source)

Given Turkish banks' eagerness to deposit gold, it is fair to say that the Reserve Option Mechanism offered banks a cheaper alternative for meeting reserve requirements than the traditional avenue of borrowing Turkish lira funds and keeping them on deposit at the CBRT. Put differently, the gold option has provided bankers with a route for reducing the size of the tax imposed on the banking sector by reserve requirements.

Gold deposit accounts

To help attract the necessary gold, Turkish banks began to offer the public the option of opening gold deposit accounts. These accounts can either be current deposit accounts, which can be withdrawn at anytime, or time deposit accounts, which are locked in for a period of time. Time deposits are remunerated. HSBC Turkey, for instance, currently pays gold depositors 1.25% for a one-year fixed deposit, as illustrated below. In effect, the tax break that the CBRT provides banks via the Reserve Option Mechanism is being shared with the Turkish public in the form of interest.

Turkish gold deposit rates (source)

The gold is held in unallocated form rather than in allocated form. When a customer deposits physical gold with their bank, the bank in turn takes its customers' grams and pushes them at the CBRT to meet a portion of its reserve requirements. A customer can also deposit Turkish lira to open a gold account. In this case, the bank can use those funds to purchase gold in order to meet its reserve requirements.  In general, banks do not seem to allow customers to withdraw their gold in physical form. ISbank is an exception, but only allows conversion of gold deposits in amounts greater than five kilograms. Turkish banks have other options for securing gold including borrowing or swapping for gold in the institutional gold market.

By mid-2013, CBRT data shows that around US$12 billion worth of gold deposit accounts had been opened by banks, representing around 200 tons of gold. But gold deposits wilted through 2015 and 2016, eventually falling to around 75 tons worth. As of April 2018, around 150 tons worth of gold deposits are outstanding, as illustrated by the blue line in the chart below.

Gold deposits at Turkish commercial banks (source)

Keep in mind that only a fraction of this 150 ton deposit base would have been opened with actual mattress gold, the majority probably being purchased by depositing Turkish lira. So although the ROM has been almost fully utilized and has encouraged the growth of gold deposit accounts, most of Turkey's 3,500 or so tonnes in mattress gold remains outside of the banking system.

Gold on the CBRT's balance sheet - What does it actually own?

Gold is currently accounted for on the CBRT's balance sheet at 571.5 million grams, or 571 tonnes, putting it twelfth on the list of countries sorted by size of gold reserves. But one needs to be careful with this amount. Most of the gold on the CBRT's balance sheet is not the property of the bank, but rather is held on behalf of commercial banks that happen to be using the ROM facility. This amount needs to be subtracted from the central bank's total holdings to arrive at the total quantity of gold that officially belongs to the Turkish state.

The chart below decomposes the quantity of gold on the CBRT's balance sheet into two categories: 1) gold being held on behalf of banks to satisfy reserve requirements (blue area), and; 2) gold held as an official reserve asset (pink area). With the ROM being introduced in early 2011, banks went from holding 0 tonnes at the CBRT to holding as much as 417 tons by late 2014, after which ROM usage fluctuated to as low as 260 tonnes before rebuilding to today's 327 tonnes. Meanwhile, the amount of gold that the CBRT actually owns clocked in at a constant 119 tonnes from 2011 to 2017. This means that (until mid-2017) all of the changes in the gold portion of the CBRT's balance sheet has been due to varying commercial bank usage of the Reserve Option Mechanism, not official changes in Turkey gold ownership.

Gold on the CBRT's balance sheet, by type

Things changed in mid-2017 when Turkey finally began to buy gold on an official basis. Since then, gold holdings have more than doubled from 119 tonnes to 244 tonnes. This reportedly comes at the behest of PM Erdoğan, who seemingly has a soft spot for the yellow metal. (Ronan Manly has written an article showing how the location of those reserves has been changing.)

As for the Reserve Option Mechanism, it was further expanded in 2016 when the CBRT allowed banks to keep an additional 5% of required reserves in the form of wrought or scrap gold. The already-existing 30% ROM tranche, introduced in 2011, had only allowed for the usage of standard gold -- gold bars in a range of weights of minimum 995/1000 fineness as defined by the Borsa Istanbul. By allowing for scrap gold - unprocessed gold below 995/1000 fineness - to be used to satisfy reserve requirements, the CBRT was hoping to draw additional gold savings out from under people's mattresses and into the banking system. The scrap gold mechanism has not been very popular. According to the CBRT's 2017 annual report, at the end of 2017, only 17% of the head-room provided by the option was being utilized.

In an attempt to attract even more gold into the banking system, Turkish authorities began to issue gold bonds and gold-backed lease certificates to the public in 2017. The one-year interest rate on these bonds was 2.4%. The first round, which ran from October 2-27, 2017, attracted just 2.5 tons, while the second round, which ran from March 26 to April 20, 2018 period, brought in 1.9 tons. I've sourced these numbers from the CBRT's May 2018 Financial Stability Report. Given the small amounts of gold that have been collected by bonds and lease certificates, these efforts can only be considered a failure.


Turkey's campaign to mop up mattress gold hasn't been very successful. Neither the original 2011 Reserve Option Mechanism nor its 2016 extension to scrap metal have attracted large amounts of physical gold from the public. The recent attempts at pushing gold bonds and lease certificates on the public have also fallen flat.

Presumably the Turkish government will have to increase its efforts if it wants to put a real dent in the 3,500 tonnes of mattress gold presumed to be held by Turks. This may require a much larger ROM gold tranche or higher interest rates on gold bonds. As it is, the 2.4% paid on gold bonds simply hasn't enticed the public into giving up their gold. But using more resources to attract gold out from mattresses will only come at the expense of Turkish taxpayers. Is it worth it? If the effort to mop up Turkey's mattress gold does expand, at some point the presumed benefits of capturing mattress gold will be overwhelmed by the costs of overcoming people's fondness for the yellow metal.

Gold and the Monetary Blockade on Iran

This blog post is a guest post on BullionStar's Blog by the renowned blogger JP Koning who will be writing about monetary economics, central banking and gold. BullionStar does not endorse or oppose the opinions presented but encourage a healthy debate.

With Donald Trump close to re-instituting economic sanctions on Iran, it's worth remembering that gold served as a tool for skirting the the last round of Iranian sanctions. If a blockade were to be re-imposed on Iran, might this role be resuscitated?

The 2010-2015 Monetary Blockade

The set of sanctions that the U.S. began placing on Iran back in 2010 can be best thought of as a monetary blockade. It relied on deputizing U.S. banks to act as snitches. Any U.S. bank that was caught providing correspondent accounts to a foreign bank that itself helped Iran engage in sanctioned activities would be fined. To avoid being penalized, U.S. banks threatened their foreign bank customers to stop enabling Iranian payments or lose their accounts. And of course the foreign banks (mostly) complied. Being cut off from the U.S. payment system would have meant losing a big chunk of business, whereas losing Iranian businesses was small fry.

One of the sanctioned activities was helping Iran to sell oil. By proving that they had significantly reduced their Iranian oil imports, large importers like Japan, Korea, Turkey, India, and China managed to secure for their banks a temporary exemption from U.S. banking sanctions. So banks could keep facilitating oil-related payments for Iran without being cut off from the dollar-based payments system. The result was that Iran's oil exports fell, but never ground to a halt. This was a fairly balanced approach. While the U.S. wanted to deprive Iran of oil revenue - which might be used to build nuclear weapons - it didn't want to force allies to do entirely without necessary crude oil.

The U.S. Iran Threat Reduction & Syria Human Rights Act of 2012 (TRA) further tightened the noose. The TRA prevented Iran from repatriating any of the oil & gas funds that were accumulating in foreign escrow accounts maintained in Turkey, Japan, and elsewhere. To enforce this restriction, any bank proven to repatriate Iranian oil money would be cut off by its U.S. correspondent bank, thus losing its connection to the U.S. banking system.

The inability to unlock funds was inconvenient for Iran. There was no useful purpose to which the pariah nation could put all the Japanese yen, Indian rupees, Chinese yuan, or Turkish lira that was piling up in its overseas escrow accounts. Iran could still buy non-sanctioned goods and services in these countries and bring them back to Iran, say food, medicine, and whatnot. But none of the oil-importing nations provided Iran with enough importable stuff that it could draw its account balances down to zero. The funds held in escrow were dead money.

Gold Enters the Scene

This is where gold was recruited as a useful payments rail for evading the sanctions. The scheme was carried out primarily through Turkey. Gold dealer Reza Zarrab, one of the most well-known expediters of the scheme, designed an escape hatch for Iranian funds along with bank officials at Halkbank, a Turkish state-owned bank. Iranian funds frozen in a Halkbank escrow accounts (denominated in Turkish lira) were transferred to accounts held at Halkbank by Zarrab's shell companies and used to buy gold in the Turkish gold market.

This technically was not illegal. While an earlier set of sanctions had prohibited banks from directly helping the Iranian government to deal in gold, it was still legal for them to help “private” persons to access gold, as long as the gold was sent to Iran and not elsewhere. Zarrab sent some of this gold directly to Iran. The rest was exported to Dubai where it was sold for dollars, the cash being returned to Turkey where Zarrab laundered it back into the banking system. Now the Iranian government could finally make use of it. Below is a chart that Zarrab made at his 2017 court trial that illustrates the complexity of the gold trade.

"Iranian gold transaction," an illustration of Zarrab and Halbank's method for moving Iran's frozen funds (source)

This size of this loophole was significant. According to Reuters, Turkish gold exports to Iran exploded from one tonne in 2011 to 125.8 tonnes in 2012, worth $6.5 billion. Another $4.6 billion was sent to Dubai, most of this ultimately destined for Iran. Below I've charted Turkish gold exports from 2008 to present. You can see the big bulge in 2012-13.

The large inflow of Turkish gold helped keep Iran afloat. It was only in mid-2013 that the "private person" loophole was finally cut off. Banks like Halkbank, which before could only be penalized for engaging in gold dealings with the Iranian government, were now prohibited from enabling gold payments for private Iranian entities as well. Gold shipments out of Turkey slowed to a trickle.

Turkish Gold Exports: 2008 to Present (millions US$)

While Zarrab continued to illegally export small amounts of gold, he focused on a different form of sanctions busting: moving Iranian funds locked in Turkey to Dubai by using fake invoices for food and medicine as cover. But this too came to an end. During a visit to Disney World with his family in 2016, Zarrab was arrested and accused of helping Iran evade U.S. sanctions. In late 2017, he accepted a plea agreement and became the key witness against a number of Halkbank officials and Turkish government officials. Only one of them, Mehmet Atilla, has been convicted to date, the rest remain at large.

The Advantage of a High Value-to-Weight Ratio

In theory, the funds held in Iran's escrow accounts at Halkbank could have been used to buy any sort of commodity, say copper, and then the copper shipped to Iran (or Dubai). The reason that gold was probably chosen as the in-between commodity is because it has a higher value-to-weight ratio than most commodities. Whereas an ounce of gold conveys $1300 in value, it would have required around 6,500 ounces of copper to convey $1300 in value. The extra bulk involved would have been very costly to ship.

Value-to-weight ratios (source)

Another advantage of gold relative to other commodities is that it is highly liquid. There are markets for the yellow metal all over the world that attract the participation of a wide range of buyers from consumers to industrial users to jewelers. Copper is much less marketable. Because Iran needed to on-sell whatever commodity was being used as the in-between commodity in order to get hard currency like euros and dollars, gold would have been the most convenient option.

A Monetary Blockade in 2018?

The monetary blockade that began in 2010 was eventually successful. Iran was forced to come to the negotiating table and in 2015 the Iran nuclear deal was signed. Zoom forward to today. Trump has left the nuclear deal and is threatening to reimpose sanctions against Iran. Could gold once again become a player in the game played between sanctioner and sanctioned?

One major difference between then and now is that there was significant global buy-in during the last round of sanctions. Prior to setting up the monetary blockade, the U.S. gained support from the United Nations Security Council, including China and Russia. So while certain loopholes were exploited, a broad consensus meant that evasion was not prevalent enough to bring the whole edifice down.

In exiting the Iran deal last month, Trump has done so alone. Presumably this lack of consensus will make any ensuing round of sanctions much more leaky than the initial ones. India, for instance, has already said that it won't comply with the sanctions because they are not UN-mandated. Now, in actuality there probably isn't much India can do to escape the sanctions. The majority of Indian banks will comply because they will be wary about the threat of being cutoff from the dollar-based payments system.

But with a wink and a nod from their national governments, a few banks who don't do much U.S. business may decide to take Iran on as a customer. They will need some sort of popular medium for expediting payments to and from Iran, say gold, euros, bitcoin, or an up-and-coming national currency such as the Chinese yuan. It remains to be seen which one would be selected, but gold certainly has a number of useful properties, as evidenced by the last round of sanctions. Whatever the case, the U.S. dollar's dominant role as a global medium of exchange can only be weakened by another round of Iran sanctions.

Sources used to write this post:

  1. Iran’s Turkish connection, Reuters, 2014 (pdf)
  2. Aggressive U.S. Monetary Policy... in Iran, 2012 (Moneyness blog)
  3. United States of America vs Reza Zarrab et al, 2017 (pdf)
  4. United States of America vs Mehmet Hakan Attila, 2018 (pdf)
  5. Iran's Gold Loophole, 2013 (pdf)
  6. Sanctions Against Iran: A Guide to Targets, Terms, and Timetables, 2015 (pdf)
  7. The Beauty of TRA Section 504, 2015 (Sanctions Law blog)
  8. OFAC FAQs: Iran Sanctions (US Treasury website)
  9. Exiting the Iran deal is a blow to financial transparency and US control, 2018 (Interfluidity blog)