Fragment has cleared more than $2M in documented @Name trades. There is no published price index. That gap is where most buyers make their mistakes — pricing by gut, by domain analogy, or not pricing at all until a competitor has already cleared the handle they needed.

This article builds the index. Two layers: intrinsic value denominated in TON, and a USD overlay that makes the model honest about the one variable it cannot fix.

Why the index has to start in TON

Every @Name on Fragment is bid and cleared in TON. The smart contract is TON-native. There is no USD settlement option. This matters for the index because a handle priced at 100,000 TON costs $182,000 at today’s price of $1.82/TON — and would have cost $270,000 at the May 9 price of $2.70/TON, and $135,000 at the early-May price of $1.35/TON. That is a $135,000 swing in USD cost for the same asset, driven entirely by TON price movement, not by any change in the handle’s market value.

A USD-denominated price index for Fragment is not wrong. It is structurally incomplete. The correct index holds TON price as a variable and presents USD figures as a function of that variable, not as a stable floor.

Component 1: Character count

Length is the most reliable predictor of clearing price in TON. The data is not subtle.

  • 2–3 characters: No publicly documented cleared 2-character handles. Three-character handles that are dictionary words or category terms (@dao at ~$410K, @nft at ~$900K) clear at institutional price. The floor is not a reflection of scarcity alone — it reflects that 3-character handles are short enough to be universal identifiers, not just usernames.
  • 4–5 characters: @boss ($500–520K), @music ($1.05M). At four and five characters, clearing price diverges significantly based on dictionary and category status. @boss clears at half of @music not because it is less scarce but because music is a higher-status universal category word.
  • 6–8 characters: Clearing prices drop sharply into the $5,000–$50,000 TON range. The exceptions are brand-exact trademark handles (e.g., a 7-character Fortune 500 brand name), which carry a legal premium that distorts the statistical baseline.
  • 9–12 characters: Commodity tier. Sub-5,000 TON with rare exceptions. Liquidity is thin — these handles clear when there is a specific buyer, not when the market bids them up.
  • 13+ characters: Near-zero clearing prices. The market does not assign material value to long descriptive handles regardless of their apparent relevance.

Practical implication: A corporate buyer seeking a defensive Telegram identity should treat the character count as non-negotiable. A 12-character handle that perfectly describes the brand will trade at commodity prices and carry no prestige signal. A 4-character handle that is adjacent to the brand but short will cost 10–50x more and appreciate with the market.

Component 2: Dictionary premium

After character count, the dictionary premium is the clearest price driver in the data. Fragment’s top trades are not short arbitrary strings — they are recognizable English words that function as universal category identifiers.

  • Top-tier English dictionary noun (music, boss, king, life, gold, code): 3–5x character-count baseline in TON. These handles attract bidders from multiple industries simultaneously, which sustains the premium across market cycles.
  • Common English word, non-category: 1.5–2x baseline. Market is more specific-buyer dependent.
  • Non-English common word: 1–1.5x baseline in most markets. Significant exceptions in Japanese, Arabic, and Malay markets where brand-register words carry culture-specific premiums.
  • Invented or arbitrary string: 0.5–0.8x baseline. The market discounts arbitrariness regardless of character count.

Component 3: Brand-adjacency premium

Brand-adjacent handles — those that match or approximate a registered trademark, Fortune 500 brand name, or dominant category term — carry a pricing premium that reflects both demand and legal risk simultaneously.

The corporate buyer reading this as “I should buy the handle that matches my trademark” is correct. The corporate buyer who concludes “the legal risk on competitor’s name means I should wait” is making an error — that risk is priced into the handle and protects the holder, not the challenger.

  • Exact-match Fortune 500 brand: 2–4x character-count baseline. The premium reflects both the brand’s marketing budget (it will eventually be sought by the brand itself) and the legal cost to challenge the holding.
  • Dominant category term (@dao, @nft, @defi): 2–3x within the relevant community’s bidding pool. Premium does not necessarily transfer outside that community.
  • Industry-adjacent descriptor: 1–1.5x. Moderate premium, moderate liquidity.

The bimodal distribution finding

Fragment’s market is not a smooth gradient from cheap to expensive. It is bimodal. The data shows two distinct clusters with thin liquidity between them.

Cluster 1 — Premium: Handles that are short (≤5 characters), dictionary or category terms, with or without brand adjacency. Clearing prices range from 50,000 TON to 600,000+ TON. At today’s $1.82/TON, this cluster ranges from approximately $91,000 to $1.1M USD. These handles have institutional buyers and auction competition. They appreciate with TON.

Cluster 2 — Commodity: Handles that are long (≥9 characters), descriptive or invented, without established category status. Clearing prices range from 0 to 2,000 TON. At today’s $1.82/TON, this cluster tops out at ~$3,640 USD. These handles have specific buyers but no market. They do not reliably appreciate.

The dead zone: Six-to-eight-character handles that are not strong dictionary words and not brand-exact trade in an illiquid middle band (2,000–50,000 TON, roughly $3,640–$91,000 at today’s price). Clearing is unpredictable. Some sit on Fragment for months without bids. This is where buyers overpay most often — they see a $30,000 ask and assume the market has validated the price, when in reality the ask reflects seller wishful pricing, not cleared trades.

The bimodal structure means that the relevant question for any potential acquisition is not “where does this handle fall on the price curve” but “which cluster does this handle belong to.” Handles in the dead zone should be treated as commodity unless there is specific evidence of institutional bidding activity.

The USD overlay — what changes with TON price

At today’s TON price of $1.82:

  • A 3-character dictionary handle with a 250,000 TON floor = $455,000 USD
  • A 4-character category-adjacent handle at 150,000 TON = $273,000 USD
  • A 6-character brand-adjacent handle at 30,000 TON = $54,600 USD
  • A 10-character descriptive handle at 500 TON = $910 USD

At $2.70/TON (three weeks ago), the same handles cost 48% more in USD with no change in TON-denominated floor. A corporate procurement process that approved a “$455K acquisition” three weeks ago and is now re-valuing at “$273K equivalent” has not seen the asset get cheaper — the TON floor has held. The USD figure moved because TON moved.

What buyers get wrong, specifically:

  1. Domain comps. The domain market prices on revenue multiples, traffic, and brand recall in a context where USD is the settlement currency. Fragment prices on TON scarcity, character count, and TON-native category status. The models share no inputs. Domain comps systematically undervalue premium @Names and have no framework for the bimodal structure.
  2. Treating USD floor as stable. The floor on a 3-character dictionary handle is 250,000 TON. The USD equivalent of that floor moves with every TON price tick. A buyer who waits three months for a “better USD price” may find the TON floor unchanged while USD costs have risen 40% from TON appreciation.
  3. Buying into the dead zone. Mid-length, non-dictionary handles at $20,000–$80,000 USD are not “affordable premium.” They are overpriced commodity. The bimodal distribution makes the middle of the market the worst risk-adjusted purchase, not the best value.
  4. Equating descriptiveness with value. A 12-character handle that perfectly describes a brand’s product category is worth approximately zero in TON-market terms. The market prices brevity and cultural recognizability, not descriptive accuracy. Corporate buyers who optimize for descriptiveness are systematically purchasing commodity handles at commodity prices while the premium market moves without them.

Reading the index as a corporate buyer

The practical output of this framework: any @Name can be scored by character count baseline (in TON), adjusted upward for dictionary status, brand adjacency, and category term weight, then converted to USD at the current TON price with an explicit note that the USD figure is a live variable.

A procurement team approving a @Name acquisition should specify approval in TON terms, not USD. “We are authorised to acquire this handle at up to 150,000 TON” is a stable mandate. “We are authorised to spend up to $273,000” is a mandate that expires the moment TON moves 10% in either direction.

The handles that appear overpriced by domain analogy are the handles worth buying. The handles that appear reasonably priced by domain analogy are in the dead zone. That inversion is what this index is for.

TON reference price used in this article: $1.82 USD (2026-05-28). Fragment Economy Intelligence.

Tomorrow: The $100K Floor — what evidence actually supports a $100,000 USD floor on premium handles, and what conditions would break it.