加密韋馱|Crypto V🇹🇭
加密韋馱|Crypto V🇹🇭|Apr 22, 2025 17:34
Last week, Space conducted daily research on short selling and even made a mediocre summary Originally, I had no connection with Huma Finance, but was unexpectedly dragged into the game by a post from @ 0xFeng that day. Coincidentally, I happened to be a practitioner of cross-border currency exchange and overseas P2P ten years ago Although opinions vary@ DrPayFi's frankness and professionalism really impressed me. Whether this industry is professional or not can be determined by just talking about it, Richard has summarized the business logic of Huma very comprehensively, and I will only add one background: What is the cross-border payment industry? Why do we need Huma Cross border payment, as the name suggests, refers to transferring money from one country to another. There are various reasons for this, such as the absence of domestic "Taobao" users overseas buying things from other countries' e-commerce platforms, European and American countries paying salaries to outsourced workers in Asia, wealthy Chinese people transferring money overseas to exchange for houses, and so on There are two types of payment service logic: -One is the payment channel: the main logic is to send and receive payment requests, that is, to send and receive orders. The principle is to help consumers and merchants keep accounts, and finally the bank settles the payment by depositing or deducting the corresponding money into or from the corresponding account Local IBPS/PROMPTPAY systems such as China, Australia, and Thailand can be accessed by scanning the code, as they are all in the same jurisdiction, making it easier to resolve any issues But if it involves countries, it's not feasible, but cross-border transactions become more complex. Different legal systems can only rely on SWIFT messages to unify transfer transactions. What was the result? Banks receive and send messages slowly, expensive, and may even receive T+5 payments. I believe all the remaining children have a deep understanding. What you may know, such as Stripe, PayPal and other payment companies commonly used in international e-commerce also belong to this category. Essentially, it is a front-end of the bank's cross-border remittance system, similar to the relationship between GMGN and RAYDIUM. The fees and the time taken for payment are in line Many PayPal merchants, due to the risk of fraud and the international credit card prosecution period, need 180 days to withdraw funds. You think it's troublesome to use, but the merchant thinks it's even more troublesome -The second type is the logic of currency exchange swap: in places like Australia and Canada, there are many currency exchange companies holding money transfer licenses (unlicensed ones are called "water houses"). Users make payments in country A, and as long as the money is issued with a receipt, the company's B country account directly disburses the funds. This is equivalent to the remittance company's accounts in different countries bearing the responsibility of settlement, rather than the bank. The advantage of this is that the payment is faster, more convenient, and users are more likely to trust it. In the past, wealthy people in China needed to use their money to buy houses in other countries, and the most important thing was to be able to receive it on the same day, because there were too many scammers. But the so-called collusion is actually equivalent to one country's account lending a sum of money as an advance to another country. If you have a small scale and only do a few hundred thousand or one million a day, it's okay to do both on your own. But as the volume increases, especially in other countries where cooperation with local three parties is required, the cost of capital occupation becomes higher Huma Finance, The solution lies in the issue of advance funding, not the remittance itself. His business logic for international remittances is simple: Can't the exchange company handle it? I'll take a few basis points from your foreign exchange business. The exchange rate in Australia is generally 2.5-3.5%, Huma's draw is completely reasonable So, where are the risks of Huma? 1. P2P black box risks: P2P black box risk: Crypto financing → USDC lending → ARF actual lending. Cannot be closed on the chain. We can only rely on compliance, which Richard has elaborated on 2. Counterparty risk: Richard said that Huma only does B2B business, which means that for remittance companies, he does not have any direct relationship with the payer or payee, but only provides USDC loans to the remittance company itself. But as those of us who have done this business before, we all know that even so, there is still counterparty risk I have also told the story of the Australian Yangtze River currency exchange being seized by the Australian Federal Police and Royal Pay running away on Space. There are also some countries with policies that have a high degree of gray scale and even unreliable rule of law, such as India in 2022, and the endless pursuit of profit oriented law enforcement in Hunan and Henan provinces. Previously, many small institutions in Australia went bankrupt, most of which were related to this Richard also mentioned that they have corresponding risk control plans for this issue, including dedicated funds loss cover, But the core solution is to only serve high rated customers, and as the borrower, strictly and continuously review the other party's credit qualifications This is certainly not a perfect solution. Once a problem arises, even if there is a legal claim, it will take a considerable amount of time and financial cost to go through the judicial process. But worrying too much about this issue is meaningless. The Huma business belongs to the broad category of accounts receivable loans, which is very mature in Europe and America and has corresponding legal remedies. Any risk control is relative, and there is no 100% foolproof solution - after all, there was a huge amount of detachment in the USDC SVB incident that year, and the losses caused are not well explained. Compared to them, this risk is much more controllable What do I think of the Huma project? From a payment perspective, it makes sense and anyone who has done business knows it From the perspective of DeFi, it is a new "bridge lending strategy": 100% available funds, unlike passive strategies such as AMM, AAVE, LSD that require redundant funds, high capital efficiency, and natural APR. And it is a U-based strategy The main issue with Huma is: The core competitiveness of the business lies in risk control and business development capabilities, but currently this set of capabilities is still off chain, and there is currently no evidence of using algorithms to replace risk control for business expansion with diminishing marginal costs like AMM or on chain lending. There is a contradiction between expansion and risk control Will the rapid expansion of stablecoin payments like USDC squeeze traditional cross-border payment business scenarios, as proposed by @ starzqeth, and subsequently affect Huma's current business In my subsequent private chat with @ DrPayFi, I also gave him two suggestions: 1. Speak less about the big company narrative of "platform application revenue" and focus more on liquidity The moat of the cryptocurrency circle has always been liquidity, whoever has a thick pool and more liquidity protocols connected wins. For example, in LST, everyone's business logic is exactly the same, and the winner depends on the liquidity of the LST token and the acceptance of the exchange. The higher these two points are, the more likely they are to generate profits in the future. And the essence of Huma is the LST of the US dollar stablecoin In addition to the business itself, it should focus on connecting all SOLs and even large liquidity agreements outside of SOLs, using HUMA's PST as the basis LP and deductible assets as much as possible. Perhaps larger Web2 may come to copy later, but liquidity barriers cannot be copied, otherwise USDT would not be the first 2. Hijacking the word 'PayFi' To be honest, if you add what you wrote to what I am writing now, most people in the cryptocurrency industry who are still in the market are unlikely to read it, and even after reading it, they may not understand it. The vast majority of them will not interact with your business, at most they will interact with your token. Game Huma, first of all, game PayFi can become something that repeatedly appears on the front page headlines and captures the attention of the cryptocurrency circle under Solana Foundation or your own or other conditions Token is a subsidiary of attention, so whenever PayFi appears, it should make everyone think of Huma. Create all events to create a conditioned reflex of Huma=Pay A Meme kernel is better than 100000 words of research and analysis
+5
Mentioned
Share To

Timeline

HotFlash

APP

X

Telegram

Facebook

Reddit

CopyLink

Hot Reads