Why Swedish Local Flex Markets Are Thin — Structural Causes

Sweden ran the most extensive local flexibility market experiments in the Nordics: four areas across CoordiNet, sthlmflex through three seasons, Effekthandel Väst, and then E.ON’s commercial scale-up to 12 SWITCH markets. The results are technically impressive and commercially disappointing. As of winter 2025/26, 6 of 11 E.ON markets had zero clearing despite bid volume present; the entire Swedish flex market activated roughly 292 MWh over a full winter season — a small industrial customer’s weekly consumption.

FlexAbility (2025) adds important aggregate data: across all active Swedish flex markets in 2024/25, activation occurred in only 6% of hours. Availability was procured 25% of hours. Registered capacity (effektförmåga) far exceeds activated volumes — Göteborg alone had 19 MW qualified in 2023/24 against a few hundred MWh of annual activation. This latent capacity finding confirms the structural diagnosis: the constraint is not physical supply, but economic and organizational barriers to activation. (Source - FlexAbility Delrapport 2 (2025))

This is not primarily a market design failure. It reflects a set of structural causes embedded in the economics, regulation, and physical grid.

Cause 1: The TSO subscription mechanism suppresses demand

The most empirically established cause. Regional DSOs buy an annual subscription (abonnemang) from Svenska kraftnät — a contracted maximum power. When they forecast an exceedance, they have two options: activate local flexibility, or request a temporary subscription increase from Svk.

Svk’s temporary subscriptions are priced at ~240–280 SEK/MWh — far below the pain threshold. When Svk routinely grants them, DSOs rationally defer to the subscription rather than run a market. Local flexibility only becomes economically necessary when the subscription is denied.

CoordiNet quantified this precisely:

Same product. Same market structure. 50× volume difference. The driver was Svk’s administrative behavior, not market design. (Source - CoordiNet D4.7.2 Swedish Demonstration (2022))

This means TSO subscription policy is a more important determinant of local flex market liquidity than anything Ei, DSOs, or market platform operators control. No market design improvement solves the problem if the TSO’s subscription mechanism makes the market unnecessary.

The implication is not that Svk’s subscription policy is wrong — granting temporary subscriptions during grid-safe periods is rational grid management. The implication is that relying on flex markets to develop under current conditions requires either (a) that Svk restricts temporary subscriptions systematically, or (b) that other congestion cost signals (e.g., under NC DR) create alternative demand drivers.

FlexAbility Delrapport 5 (2025) adds a cross-cutting finding: 50% of active flexibility providers (workshop participants) report increased interest in participating in Nord Pool wholesale markets as an alternative to local flex markets. When the national market offers better value than local flex markets, supply migrates away — further thinning local market liquidity and increasing concentration risk.

Cause 2: A sällanköpsmarknad — rare activation destroys the FSP business case

Even in active markets, activation is rare and unpredictable. Winter temperatures vary by 3× in activation frequency between mild and cold years. A market that clears 300 MWh one winter and 30 MWh the next cannot support the fixed costs of qualification, system integration, and standby readiness.

The phrase sällanköpsmarknad — “infrequent-purchase market” — emerged from DSO operator interviews in Ei’s 2023 evaluation and captures the core problem: flexibility providers cannot cover their costs on activation revenue alone when activation is weather-dependent.

The solution: availability contracts (fast bid agreements) that pay a capacity fee regardless of activation. CoordiNet piloted these from winter 2020/21 onward. sthlmflex introduced ShortFlex Availability in season 3. E.ON’s SWITCH markets use Säsongstillgänglighet (ST) products as the primary market tier. (Source - sthlmflex säsong 3 (2022-2023))

The limitation: availability contracts don’t fully solve the problem; they just redistribute it. The DSO now bears the cost of availability payments in years with mild winters and low activation — potentially paying for capacity that is never used. This is rational from a system reliability perspective (the option value of having the resource available is real) but requires DSOs to accept weather-induced cost variability, which creates accounting and procurement planning challenges.

Bottom line: the seasonal availability payment is a necessary but not sufficient fix. It makes the FSP business case viable; it does not make the DSO business case compelling when Svk subscriptions provide an escape route.

Cause 3: Energy price spikes destroy cheap supply at the worst time

Energy prices and local flex market liquidity move inversely — exactly when congestion is highest (cold, high-load periods), energy prices are also highest, which destroys the cheapest flexibility supply.

Mechanism: district heating companies with electric boilers are the backbone of large-capacity, low-cost flexibility in most Swedish markets. They provide cheap flexibility when they’re running electric boilers (i.e., when electricity is cheap relative to their alternative fuel cost). When electricity prices triple (as they did in winter 2021/22 in SE3), district heating companies stop running electric boilers and switch to alternative heat sources. Their flexibility disappears from the market just when the DSO most needs it.

sthlmflex Season 2 (2021/22): the coldest winter in the dataset (highest activation need) coincided with tripled SE3 electricity prices → the market had its lowest cheap-bid liquidity. 878 MWh activated at 883 SEK/MWh average, against 9,965 MWh at 248 SEK/MWh for CoordiNet Uppland under more favorable conditions. (Source - CoordiNet D4.7.2 Swedish Demonstration (2022), Source - sthlmflex säsong 3 (2022-2023))

This is a structural market design problem: the production function of large thermal assets means they substitute between flexibility and their primary energy product based on electricity prices. No flexibility market design can eliminate this correlation unless it either (a) secures their capacity under long-term availability contracts that lock in the flexibility revenue stream regardless of spot prices, or (b) develops alternative supply from resources that do not substitute with electricity (EVs, dedicated batteries, industrial processes with stable operating schedules).

Cause 4: Concentration risk — a few large actors dominate supply

Sweco’s 2025 structural analysis reveals a consistent pattern across all Swedish markets:

Aggregators contribute market diversity — many small resources, low per-unit capacity — but energy companies provide depth. In cold winters, a single district heating company with a 50–200 MW electric boiler determines whether a market clears at all.

This creates a market structure risk: if two or three large energy companies exit a local market (because Svk’s balancing market pays better, or because they’ve closed or repowered their boilers), the market loses most of its clearing capacity. No aggregated household portfolio currently substitutes for the large thermal asset. (Source - Sweco Kartläggning av lokala flexibilitetsmarknader (Ei, 2025))

The longer-term picture: as EV fleets, large battery storage, and heat pump aggregation scale up, the concentration risk should ease. But in the current market, this structural fragility is a primary reason DSOs prefer Villkorade Avtal as the backstop — villkorade avtal deliver close to 100% of contracted capacity; flexibility markets deliver ~80%, and the shortfall is concentrated in exactly the high-need moments when energy prices are high.

Cause 5: CAPEX bias — DSOs don’t benefit from procuring flexibility

The Swedish revenue regulation (intäktsramsreglering) until 2027 contains a structural incentive that makes grid investment more attractive to DSOs than flexibility procurement:

• Capital investment earns a regulated return on the asset base (WACC × RAB) over the asset’s lifetime — typically 40+ years at the allowed WACC rate
• Flexibility procurement is operating expenditure — recovered at actual cost, but earns no regulated return

The formal economic mechanism underlying this asymmetry is the Averch-Johnson effect: under capital-return-based regulation, a monopolist has a systematic incentive to overinvest in capital because the allowed return is tied to the asset base. The incentive is not a conscious DSO choice; it is structurally embedded in the revenue cap’s treatment of capital versus operating costs. (Source - Konsultrapport Kostnadsincitament Regionnät Termicum (2025))

The §9.5.8 mechanism — what the regulation actually says: Ei’s revenue cap Handbok (§9.5.8, EIFS 2023:5) classifies flexibility service costs (kostnader för flexibilitetstjänster) as opåverkbar opex — uncontrollable operating costs. This means:

• Flex procurement costs are recovered at 1:1 actual cost — no efficiency deduction applied
• Ei cannot reduce a DSO’s revenue cap because the DSO paid market rates for flexibility
• The legal basis is ellagen 5 kap 12a§ (2022) and Art. 32 Dir. 2019/944
• Condition: only applies to products approved by Ei under Förordning 2022:585 §11. First Ei-approved products were published December 2025 (LFM-h, LFM-p, LFM-e)

This is a more nuanced picture than “efficiency penalties on flex”: the efficiency penalty framing is not accurate. Flex costs are not penalized — they are simply not rewarded. The DSO recovers what it spends but earns no return on the capital equivalent of that expenditure. The cable it could have built instead earns a return at WACC for four decades. (Source - Ei Handbok Intäktsram 2024-2027 (2023))

What does not fix the bias: The ongoing transition from kapacitetsbevarande (replacement cost RAV) to förmögenhetsbevarande (historical cost RAV) — recommended for RP5 by DFC Economics and Montell och Partners — does not address the CAPEX bias. Both valuation methods treat the capital/opex asymmetry identically: capital earns a return, opex does not. The RAV transition primarily changes how assets are valued and what inflation risk DSOs bear; it has no effect on the incentive to prefer build over buy. The separate reform that does address the bias is the TOTEX/lösningsneutralitet mechanism. (Source - Konsultrapport Return on Investment Elnät DFC Economics (2024))

A DSO that resolves congestion by buying 50 MW of flexibility for SEK 5M/year earns no regulated return on that expenditure. The same DSO that builds a cable costing SEK 100M earns a return for 40+ years. FlexAbility (2025) makes this concrete: in the bus depot case study, the economically correct answer (flexibility at 17:1 NPV advantage over grid investment) loses to the regulatory-correct answer under current regulation — CAPEX earns a return for 40+ years while OPEX earns nothing. Full case study at Congestion Management › Quantified economic value of DSO flexibility procurement. (Source - FlexAbility Delrapport 3 (2025))

Ei‘s RP5 (2028–2031) TOTEX reform introduces lösningsneutralitet — solution neutrality — through TOTEX benchmarking: a DSO that handles congestion through flexibility procurement scores identically to one that builds grid. This directly addresses the bias. Full regulatory effect from 2028. (Source - Ei Inriktning intäktsramar 2028-2031 (2025)) One nuance added by FlexAbility: TOTEX benchmarking must not penalize proactive grid investment — building capacity ahead of demand to meet future electrification needs. Pure cost efficiency scoring could discourage anticipatory investments that the energy transition requires. (Source - FlexAbility Delrapport 3 (2025))

Until 2028, the CAPEX bias remains operative. This does not mean DSOs uniformly prefer grid investment — Ei’s monitoring shows flexibility market development — but it means the incentive structure does not support flexibility as a rational DSO cost strategy.

FlexAbility Delrapport 5 (2025) provides decisive interview confirmation: every DSO interviewed cites intäktsreglering as a barrier. DSOs running flex markets describe the procurement cost as “välgörenhet” (charity). This is not a fringe complaint — it is the universal experience of market participants. The CAPEX bias analysis from FlexAbility DR3 and Sweco is now supported by cross-industry qualitative evidence from all DSO types. Accordingly, intäktsreglering ranked #1 barrier in the FlexAbility workshop (38 points — ahead of all other barrier categories combined). (Source - FlexAbility Delrapport 5 (2025))

Cause 6: Markets close when grid investment resolves the bottleneck

Several Swedish markets have closed not because they failed but because they succeeded in a narrow sense: they bridged the gap until grid reinforcement arrived. This is the intended dynamics — flexibility as a bridge, not a permanent substitute.

Södra Skåne (the dominant Swedish market by price and activation volume): E.ON’s Skåne flex market addressed congestion on the Söderåsen-Barsebäck-Sege 400 kV corridor. In October 2024, the third sequential upgrade to this corridor was completed, adding 600 MW — roughly Malmö’s entire cold-day demand. The market became “much less necessary or even unnecessary” for the primary bottleneck. Fill rate dropped from 58.8% (V2023/24) to 13.2% (V2025/26); zero clearing in most sub-areas despite bid volume. (Source - SWITCH Marknadsdata (info.switchmarket.se, 2026))

sthlmflex (closed): a warm 2022/23 winter + Stockholm Exergi running 200–300 MW of local CHP + Svk granting temporary subscriptions → no structural congestion to solve. The market had no demand to exist.

This dynamic is not a problem per se — it confirms the flex market performed its intended function. But it means Swedish flex markets are geographically unstable: they form at current grid bottlenecks, and dissolve when those bottlenecks are resolved. A market ecosystem requires persistent, widespread congestion — which is structurally inconsistent with a well-functioning grid.

The structural implication: as Svk completes NordSyd and DSOs execute their grid investment programs, current flex market areas will see reduced congestion. New congestion areas (driven by electrification of transport and industry) will emerge — but there is no guarantee they will be the same geographies or involve the same actors.

Cause 7: Regulatory uncertainty has deferred investment in the market ecosystem

Several market actors — aggregators, platform operators, FSPs — have deferred investment in flex market infrastructure because the regulatory framework has been “forthcoming” for years:

• The NC DR was announced in 2022, expected in force 2024, now expected 2026. Each delay is another year where investing in a platform, qualification infrastructure, or aggregation technology carries regulatory risk.
• Ei’s CAPEX bias reform was flagged as a problem in 2023; the fix comes in 2028. DSOs in 2024–2027 still face the old incentives.
• Villkorade avtal’s regulatory status (Ei2025:01) was only clarified in 2025. DSOs that signed agreements without approved methods may need to retroactively address compliance.

Sweco’s 2025 interviews document this explicitly: “lack of regulatory clarity” was the most commonly cited barrier to flex market development by all actor types. (Source - Sweco Kartläggning av lokala flexibilitetsmarknader (Ei, 2025))

Cause 8: The thin market problem — insufficient demand meets fragmented supply

Even after controlling for the above causes, thin markets exhibit price-floor dynamics that indicate insufficient competition. The V2025/26 SWITCH data shows price uniformity at 3,000–3,500 SEK/MWh across 10 of 11 markets — with zero clearing in 6 of 11 despite bids present. This suggests either a coordination floor (FSPs coordinating minimum acceptable prices) or a genuine demand floor (DSOs not willing to pay more than their subscription alternative).

In six markets: DSOs apparently found the subscription alternative cheaper than clearing at 3,000–3,500 SEK/MWh. In four markets: DSOs did buy, but at very low fill rates. Only Bålsta showed a meaningful fill rate (22.7%) — indicating a genuine localized structural constraint that the market is successfully addressing.

This is a thin market pathology: too few buyers with too low willingness to pay, combined with too few sellers with too high reserve prices. The market cannot clear at any price that makes both sides better off simultaneously. This is structurally different from a well-functioning market with a temporary demand shortfall. (Source - SWITCH Marknadsdata (info.switchmarket.se, 2026))

The interaction effects

These causes are not independent. They reinforce each other:

• The TSO subscription (Cause 1) reduces demand → rare activation (Cause 2) → FSP supply thins → concentration risk (Cause 4) worsens → thin market (Cause 8)
• CAPEX bias (Cause 5) → DSOs prefer grid investment → markets lack committed demand → rare activation (Cause 2) → FSP business case deteriorates
• Regulatory uncertainty (Cause 7) → aggregators don’t invest → small-resource supply stays thin → concentration in large thermal assets (Cause 4) → market fragility

The market is caught in a self-reinforcing low-equilibrium: low demand (because grid alternatives exist), low supply diversity (because aggregators haven’t invested), low clearing rates (thin market), low FSP confidence (because rare activation is unpredictable), low DSO commitment (because the backstop works reliably).

What could change the equilibrium

Each of the structural causes has a corresponding intervention:

No single lever solves the problem. The TOTEX reform (2028) is the highest-leverage regulatory change because it aligns DSO incentives with the market-first principle for the first time. The NC DR’s formal demand signal — derogations for non-market procurement that are time-limited and publicly reported — creates accountability for market development. Together these two changes should shift the equilibrium. But the transition period (2026–2028) is likely to see continued thin markets.

A new supply-awareness lever — EIFS 2026:8 §4.5 (2027): From 1 January 2027, a DSO that trades flexibility services on a local market must inform every user whose facility lies in the trading area — on mina sidor — that the market exists, name it, and explain how to participate (while staying neutral on specific providers). The konsekvensutredning’s rationale is explicitly liquidity-oriented: these markets “bekostas av nätföretagens kunder och kräver ett högt deltagande för att bli samhällsekonomiskt effektiva.” This is the first regulatory instrument that directly attacks the demand-side awareness dimension of the thin-market problem — most potential FSPs simply don’t know the market exists. It is only a partial lever (awareness ≠ a viable business case; the activation-frequency and energy-price-correlation causes are untouched), but it could meaningfully widen the funnel of candidate resources, especially household/aggregated supply, alongside the consumer-DR information duties in the same föreskrift. (Source - EIFS 2026-8 Nätföretags Information till Elanvändare (2026))

The DSO toolkit problem: only 15% consider joining a flex market

FlexAbility (2025) analyzed Ei’s compilation of DNDP submissions (Ei PM2025:03) and found that most DSOs are not engaging with the market tools available:

• ~50% use no flexibility services at all today
• ~20% (all large DSOs) use flex services or villkorade avtal
• ~40% mention villkorade avtal or bilateral agreements
• Only ~15% consider creating or joining a local flex market
• Only 5 DSOs have used formal redispatching

The dominant tool by far: effekttariffer — the price signal that most DSOs planned to implement by January 2027 under EIFS 2022:1. Note: EIFS 2022:1 is being repealed by June 2026 (before the deadline); effektavgifter remain permissible but are no longer mandatory (Source - Ei Effektavgifter webb (2026)). This reveals that the flex market development story primarily applies to the large DSOs (E.ON, Vattenfall, Göteborg Energi, Ellevio). The 155-company DSO landscape is largely untouched by market-based flexibility. (Source - FlexAbility Delrapport 2 (2025))

Why the current thin-market state is still progress

A final calibration: thin markets, in the Swedish context, are not failures. They represent accumulated capability:

• DSOs have qualified resources, built market-operator relationships, and learned what works operationally
• Platform operators (SWITCH, NODES) have developed market infrastructure that can scale
• FSPs have investment in monitoring, communication, and settlement systems
• Grid prequalification processes (informal) exist and are practiced
• Ei has a baseline dataset for flexibility market supervision

The question is not whether this capability will be needed — electrification demand growth makes local congestion management essential. The question is whether the regulatory and market changes underway will arrive before the capability investment atrophies from lack of activation.

Related pages

• Flexibility Market — Swedish market landscape and pilot history
• Congestion Management — subscription mechanism mechanics; SE4/SE3 price spread
• Villkorade Avtal — the backstop mechanism and why DSOs prefer it
• Balancing Markets — FCR-D as the competing option for battery operators
• Aggregation — actor composition problem; concentration risk
• Ei — CAPEX bias and RP5 TOTEX reform
• NC DR Implementation in Sweden — What Changes and for Whom — what regulatory changes are coming and when