Finance is full of misconceptions. TradFi isn’t, for example, very centralized. The global banking system is incredibly decentralized, with all different kinds of national and regional authorities and entities tying it together.

Another myth is that old technology is costly and causing people to not have access to banking services. This isn’t true either. Often, high prices or low levels of access can be traced back to cumbersome regulations or a lack of competition. Mainframe transaction processing is cheap and efficient and those “legacy” mainframes are mostly just a few years old, even if the code they’re running contains elements written decades ago.

For these reasons, I have long believed that blockchains may struggle to replace credit cards, debit cards, and traditional bank accounts. Older technologies have the advantage of excellent performance, lower costs, and large installed user bases. However, what banking and traditional finance are missing, might be the extreme competition that is arriving from the blockchain sector.

In theory, blockchains should NEVER be more efficient than centralized systems for simple transaction processing like payments. A blockchain ends up copying payment data across many nodes while a centralized system processes the transaction just once. However, a similar kind of logic applies to telephone calls. The internet is just about the worst possible way to set up a telephone or video call. Packet networks without quality of service (aka the internet) are just plain unreliable. Voice over IP calls and video calls have much lower quality and are less reliable than traditional circuit-switched phone calls.

Despite the internet’s inherent drawbacks, it has come to dominate communications because it is cheap, widely available and internet communications services are intensely competitive.

The same factors may be starting to shape the future of finance. Blockchains may be, technically, less efficient than centralized transaction processing, but the intense competition and innovation are both driving down costs and adding features at a pace that legacy service providers do not seem to be able to match.

Blockchain transaction prices are already consistently below those of traditional banking providers. Blockchains like Solana, Aptos and multiple Ethereum Layer 2 networks are shaving costs per transaction to well below a penny. Internal research conducted at EY shows that, if we were able to move all of our own internal financial transactions on-chain, we could save more than $100 million annually.

So why aren’t companies and individuals moving their transactions on-chain, then? They are, but slowly. An entire ecosystem of financial services is emerging on-chain, including companies like Franklin that can handle payroll in both on-chain and off-chain environments, and at lower costs than traditional providers. The effects of inertia are even larger in the business space than with consumers, and blockchains still lack key features like robust levels of regulatory compliance, mature fraud prevention analytics, and importantly, privacy.

These gaps are not deal breakers; they are just speed bumps. I get pitched on one blockchain security solution a week and many smaller banks are eyeing deep integration with on-chain finance as a key differentiator. The programmability of blockchains means that missing features are easy to add, and Moore’s Law means that implementing those features gets cheaper every year.

For the traditional finance ecosystem, the coming decade is critical for shaping the finance sector of the next century. Banks should heed the fate of telecom giants prior to the explosion of the internet and mobile technologies; virtually none retained their original form or structure, even if some brand names still exist.

How much time? In the United States, the entire cycle of deregulation, monopoly break up, reconfiguration, and shift to internet and mobile took place over 20 years. Please set your timer.

Disclaimer: These are the personal views of the author and do not represent the views of EY.

Note: The views expressed in this column are those of the author and do not necessarily reflect those of CoinDesk, Inc. or its owners and affiliates.