Understanding the friend.tech economic model with mathematics: Finding the optimal solution for the SocialFi price curve

Venture capital firm ABCDE researcher Loki delves into the economic model of the popular social app friend.tech, analyzing the similarities and differences of competitors Cipher, PostTech, and NewBitcoinCity, discussing what kind of price curve is needed for a successful SocialFi. Here is the content of ABCDE's report:

1. Comparison of Pricing Curves & the Cost of Changing Slopes

Since October, the competitive landscape of SocialFi has gradually become clearer, with some competitors gradually fading from the market view. Looking back at the development process of friend.tech (referred to as FT below), the economic model, especially the pricing curve, has played a very crucial role. Specifically, the pricing curve of FT has the following characteristics:

• The positive and negative differences ensure that as the number of people increases, the price continues to rise, and the rise becomes faster and faster, ensuring that early participants make money;
• 16000 achieves a relatively reasonable community size carrying capacity;
• As the number of people increases (especially after 100-200), the curve becomes steeper, price fluctuations are higher, and the carrying capacity gradually weakens;
• The leftmost part of the curve is the most lucrative buying range at present, but this part of FT is monopolized by Bots, constituting a form of income similar to "MEV."

For more detailed explanations, please refer to: Detailed Explanation of the friend.tech Economic Model: Game Theory, Expected Value, and Demand Curve Illusions

As for competitors, Cipher, PostTech, and NewBitcoinCity have completely retained the FT formula. All protocols are still constructed in the form of quadratic functions, retaining the characteristics of first derivative > 0; second derivative > 0; third derivative = 0, which will ensure that the FOMO/money-making effect of FT continues to exist.

The curve changes in New Bitcoin City mainly come from changes in the valuation currency and BTC price, while SA and TOMO have made some adjustments to the form of the curve. SA added a linear term and a constant term on the basis of the quadratic term (K²), and reduced the coefficient of the linear term. This change theoretically makes the curve overall smoother, slower to rise, and the initial price increases. However, due to the very small value of the constant term in SA, this change is not easily noticeable. TOMO's change is simpler, just reducing the coefficient of the quadratic term by about 73%.

It can be seen that SA and TOMO essentially change the growth rate of the curve. Based on this change, under the same key supply quantity, the prices of SA and TOMO will be lower, with SA's price level roughly between 15%-20% of FT, and TOMO's price at 37% of FT.

Overall, this change is not very innovative. A smoother price curve is a double-edged sword for imitators. On the one hand, FT provides a benchmark value, and it is reasonable for the price on the imitator to be lower than FT itself, lower prices will bring better acceptance and greater user capacity. But on the other hand, a smoother curve means a poorer wealth effect, which is one of the key factors that attracted hundreds of thousands of users to FT.

Of course, a steep price curve is not without its costs. The other side of the spiral rise is the spiral decline. In the past week, the TVL of frien.tech has decreased from 27,000 ETH to 21,000 ETH, a decrease of less than 20%, but the resulting price collapse and the 33 betrayal have far more consequences.

2. FT's Gray Rhino: Net Capital Outflow

The issues of FT's Bots and high transaction fees are visible to everyone, and the net capital outflow they bring is killing friend.tech. As shown in the figure below, the TVL of friend.tech all comes from user deposits, where user transactions produce PnL and royalties earned by the issuer if not withdrawn, but reinvested, this money will remain within the protocol. However, the "MEV income" earned by Bots and the fees earned by the protocol will directly become net capital outflow.

The "MEV income" earned by Bots is difficult to quantify, but the entry of DWF founder AG into FT in September is a typical case. The first buy-in price displayed on the FT front end is 0.4 ETH, which means that the Bot directly bought more than 80 Keys, with an average price of 0.135 E. These Keys were then sold out over the next 48 hours, at prices ranging from 1.1E to 1.5E. Based on this estimate, the Bot in AG's Room earned about 100 ETH in total, and all these profits came from user losses.

The transaction fee part is more quantifiable. According to Dune data, as of October 25, the cumulative transaction fees attributable to the project side were 13,840 ETH. Calculated based on the TVL peak of 27,000 ETH, users have deposited at least 40,000 ETH. Even without considering Bot MEV income, KOL royalty net withdrawals, and net outflows from fake account scams, FT has already taken away over 30% of users' capital in just three months.

When TVL rises, users' feelings are not so strong. But once TVL enters a downtrend or even just sideways, the impact becomes unusually strong. Protocol drainage, Bot MEV income + KOL royalty net withdrawals + net outflows from fake account scams are all brought about by 【non-trading】, if we estimate the latter three based on a value of 5000 ETH, which is already very conservative, then the total cumulative user deposits would be 45,000 ETH.

As mentioned in a previous article, the nominal value of Key is roughly three times the real TVL. So when the TVL is 27,000 ETH, the nominal value of all Keys is about 81,000 ETH. Compared to the 45,000 ETH of capital invested, users have received an average positive return of 80%, and when the TVL drops to 21,000 ETH, the nominal value of all Keys drops to 63,000 ETH, and users' average returns drop to 40%. It can be seen that the nominal return on Key is leveraged. If the TVL continues to decline to 15,000E, the total nominal value of users will be equal to the total invested capital, and considering transaction costs and bid-ask spreads, users will enter an overall loss state.

The breakdown of FT's 33 consensus has already begun to trend towards Tomo. If the high protocol + Bot drainage continues, the collapse of FT and other SocialFi platforms is only a matter of time, and with the decline in nominal returns, the collapse trend will accelerate. We had hoped that friend.tech would address the issues of protocol drainage and Bots, but it seems that no changes have occurred. And the recent changes in the point rules have objectively led to trading by users seeking points, further increasing transaction friction; and founder 0xRacer has also withdrawn a large amount of fees earned by his Key.

3. How Can the Curve Be Further Improved?

Furthermore, if we still maintain the premise of P = K²/C+ D (where C and D are constants), setting the pricing formula needs to consider the following factors:

Curve Growth Rate & Price

The faster the growth rate, the more FOMO, mainly achieved by increasing the constant C. Competitors generally lower the growth rate to make the curve smoother. However, the main point of this approach is to maintain the "low price" of Key, as the TVL of imitators will be significantly different from FT, so having a lower price for the same holder is more reasonable.

Number of People the Community Can Accommodate

The growth rate of the curve also determines the upper limit of the number of community members. If a higher number of members need to be accommodated, the curve needs to be smoother:

(1) Increase the constant C

(2) Set up segmented functions, with later segments smoother

(3) Calculate the proportional relationship between P and FT-Key's P under the same X conditions

"MEV Value" at the Left End of the Curve

Address the issue of being exploited by Bots' "MEV"

(1) Increase the intercept term D so that the initial price > 0 (Tomo set D but the value is very low and can be ignored). This approach also has its drawbacks: it leads to a decrease in the multiple of the wealth effect

(2) Add a smoother or horizontal curve at the left end

(3) Fixed price IDO (pre-sale system, different from 2 in that one is first come first served, and the other is a fair sale)

(4) Allow homeowners to pre-purchase

From the shape of the curve, there are two types of improvement ideas, one is directly changing the parameters of C and D. This is also the most common way of improvement currently, and changing the constant D can also solve the MEV problem to a certain extent.

The second form is to set segmented functions. This approach can set different parameters within different price range intervals to achieve different purposes. For example, setting a smoother or even horizontal curve in the first half of the curve to resist MEV or initiate a form of IDO. The IDO model is more prominent on Tomo in solving Bot MEV and failed issuances, showing positive significance.

There are costs associated with this as well. If a flat curve is adopted on the left end, the initial wealth effect will be greatly reduced, and the quantity of supply at the left end needs to be further considered, as an oversupply may consume potential buying or wealth effects.

4. Beyond KOLs, What Else Can Keys Carry?

An objective fact is that the "services" or "information" provided by most Room Owners are not enough to support the value of Key, or rather, the price of Key is generally overestimated. The reason for this problem is that the speculative demand and point-scoring demand of friend.tech have confused the real utility demand, and FT and imitators have made choices based on business purposes.

Most people only see Key as a social token, but in fact, Key can represent any asset. friend.tech has given us an idea: to introduce the issuance and trading of assets as "Fi" into "Social" to complete the final loop of SocialFi. For FT and most imitators, Key represents the personal brand or personal reputation of KOLs, but this does not mean that SocialFi is like this. Even if still based on FT, any asset can be loaded into Key, such as equity or tokens of Web3 projects - some have already done this. In this case, Key represents Tokens or shares; or by using FT to complete an IDO, Key represents investment shares or future claims (perhaps some projects will do this soon).

It currently appears that the functions of FT and imitators are too simple and do not meet some derivative needs well. Another approach is to introduce 【asset issuance】 into existing Web3 social/content products like DeBox, CrossSpace, etc. For example, DeBox, positioned as the most native DAO governance platform, has already built a social platform covering chat, dynamics, community functions based on DID, and provides voting, proposals, Token authorization detection, and transaction components.

With a large number of users, strong social connections, information, management tools, and trading tools, DeBox currently has 1.5 million registered users, over 100 million daily messages, and has high scalability in terms of functions, making it suitable for introducing an effective asset issuance solution and an economic model and pricing curve compatible with the business type.

Here, assets include but are not limited to specific content, decentralized groups, or even MEMEs with no real meaning but with collective will; then, a series of social tools and infrastructure provide services for these assets, and the value of Key will truly complete the loop.

Ultimately, the biggest difference between Fi and Ponzi is whether the assets exist and have value, and we can never ignore this fact.