XRP Blog

How much Ripple would MoneyGram need for a full scale implementation?

How Much XRP will MoneyGram Need to Implement xRipple Worldwide

Top Hardware Wallets
How much XRP does MoneyGram need to hold at any given time if they decide to roll out xRapid for all international money transfers? Good question! I’ll dig in and provide my estimate in this article. First off, the xRapid site is not a treasure trove of information. Ripple calls it Low-Cost Liquidity and describes it as

xRapid is for payment providers and other financial institutions who want to minimize liquidity costs while improving their customer experience. Because payments into emerging markets often require pre-funded local currency accounts around the world, liquidity costs are high. xRapid dramatically lowers the capital requirements for liquidity.

Here is a somewhat generic diagram Ripple uses to show how xRapid works:xRapid

On a use case page, Ripple shares the following key features of xRapid: Instant settlement, No need for nostro accounts, and Competitive FX rates.  A nostro account refers to an account that a bank holds in a foreign currency in another bank. FX refers to Forex, the world’s largest most liquid market where currencies are traded. Ripple states no need for nostro accounts, not a reduced need for nostro account balances. They state competitive FX rates, but I couldn’t find any information on why a xRapid user would get competitive rates and what kinds of discounts would be normal.

If the Ripple site was designed to give you a firm understanding of how much Ripple would be needed for a rollout of xRapid, even in general terms, I couldn’t figure that out. In a prior blog post, I took a stab at estimating that a switch to xRapid would save MoneyGram about $9 million per year, an extremely significant amount of money. I backed into that number based on estimates I found reasonable. I’ll do the same and walk through how much XRP would be needed by MoneyGram.

MoneyGram Agents are Independent

First off, we have to understand that MoneyGram agents are independent. You can sign up to become a MoneyGram agent if you have a business that could add MoneyGram’s remittance service to your mix of services. They have a bunch of FAQs for potential agents. MoneyGram does offer direct services via their apps and websites but if we are to look at their claimed 350,000 agents, we have to assume these are independent businesses. Being independent, MoneyGram can force them to use xRapid with a model where they held XRP, or create marketplace incentives where agents would be incentivized to use XRP and make their own decisions. Or MoneyGram can just centralize the use of XRP and agents do not need to hold XRP or transact in XRP.

Unlikely Scenario – All Agents and MoneyGram hold Ripple – $8.3 Billion in Ripple Needed

This is a crazy example I know but let’s assume this is the model and that MoneyGram can educate and help every agent buy and sell the amount of Ripple needed to make international transfers work. In this scenario, if I were an agent I would hold a supply of XRP and then the transactions could be done completely in Ripple, based on market prices. For a $100 transfer, I would accept $100 in my fiat currency and then I would send MoneyGram $100 worth in Ripple along with the MoneyGram Ripple transaction fee. MoneyGram could then send the XRP to the pickup location when the recipient picks up the cash requiring the pickup location to receive XRP but then hand out fiat currency.

If we assume this becomes the final model, my estimate is every agent will want to hold at least 4 days of average daily transaction volume to ensure they have enough Ripple on hand to complete transactions over weekends, holidays, and general market access challenges. MoneyGram’s annual transaction volume is roughly $25 billion (had $50 billion initially and corrected all figures that referenced this number on 1/18/18 at 9:45pm HST). With 350,000 agents, that averages out to $70,000 or so per agent per year. So, each agent would need about $800 in Ripple to have 4 days worth of average daily transaction volume on hand. This works out to $280,000,000 of Ripple needed by the agents on any given day. MoneyGram would also need to hold Ripple. My guess is a two-month volume stockpile would be a healthy balance or roughly $8 billion in Ripple.

In this scenario, the MoneyGram ecosystem would need to hold about $8.3 billion in Ripple at all times. At the current market cap, this equates to roughly 13% of the entire market capitalization. That’s a ridiculous percentage of the market cap so if my estimates are correct, the market cap of Ripple would have to rise to make room for this demand.

The pros and cons of this unlikely scenario are obvious. For one, MoneyGram and agents take on the risk that XRP’s value will fluctuate. Agents would also have to have low cost, fluid, near real-time conversion services to convert XRP into their fiat currencies. After all, MoneyGram customers are not transacting in XRP. This is not the model they are using, I am sure, but I ran the numbers to just get a sense of the XRP needs in this scenario.

Likely Scenario – MoneyGram centralizes Ripple-based exchanges – $4 Billion in Ripple

In this scenario, I am less clear in how centralized management would work. As far as I understand the model, MoneyGram agents would not have to change what they do or how they do it. They would still be accepting fiat currency, giving out fiat currency, and getting credited/debited in fiat currency. MoneyGram would be changing their internal process for international money transfers using xRapid.

The theory is MoneyGram would save money by not needing a single nostro account, according to their use case page. So with no nostro accounts, we have to assume that MoneyGram would need access to local financial institutions that can convert XRP to fiat for every one of their markets, daily, as they settle transactions with agents the very next business day.

When will this financial institution infrastructure converting XRP to fiat for MoneyGram exist? Will they charge fewer fees to make this transaction than the current fees MoneyGram pays for using existing infrastructure? I’d love to see a post from the MoneyGram CFO that adds clarity on how much money is saved and the process used to convert to fiat. If XRP is used, how do these financial institutions and MoneyGram reduce XRP price fluctuation risks?

If this is indeed viable, I would estimate that MoneyGram would need to hold about a month of XRP to have an adequate supply on hand, or about $4 billion of XRP, or roughly 6% of the current market cap. If this indeed is correct, the price of XRP will need to rise to accommodate the supply and demand of MoneyGram transactions.

The Missing Link – Financial Institutions for Ripple to Fiat Liquidity

Even if Ripple increases transaction speed and reduces transaction costs, at the end of the day MoneyGram customers are sending and receiving fiat currencies, not XRP. So with both of these models, and there are other models, I still don’t see how MoneyGram is going to go from XRP to fiat in over 200 countries without a large network of financial institutions that can provide that conversion service, at scale in as real-time as needed. Otherwise, MoneyGram is still going to need to hold significant amounts of fiat currency in all of the markets they currently support. Will Ripple become this fiat to XRP marketplace? Or at least get it started in key markets?

What’s Missing?

My quick analysis is MoneyGram would need to hold about $4 billion worth of XRP at all times. I see this working if MoneyGram has access to fluid, real-time XRP to fiat conversion financial institutions in the markets they serve, which is not yet the case. And that conversion service fees need to be lower than what they are currently paying. I get it that xRapid works and it indeed can lower the transaction costs. But MoneyGram’s business is more than just a conversion and I am looking forward to seeing more technical information on the entire chain of processes that MoneyGram needs to implement for a full-scale XRP-based system is adopted. Very promising! But still a bit cloudy.

On a short side-note, I wonder what the folks at xRapid Medical think about Ripple using the xRapid name.


  1. Your analysis seems apt, but you may be using an incorrect figure for MoneyGram’s annual transfer volume. You write:
    “MoneyGram’s annual transaction volume is roughly $50 billion”

    But the article you linked to says:
    “Since the late 80s, Western Union’s remittances business has always been 3-4 times larger than MoneyGram’s, currently processing about $50 billion more in annual transfers”

    The graph below this statement shows that MoneyGram processes about $25 Billion in annual transfers. Just wanted to point that out!

    1. I updated the figures. The change didn’t seem to make too much of an impact on the unlikely scenario amount of Ripple needed. Appreciate the correction!

Leave a Reply

Your email address will not be published. Required fields are marked *