Bund desks should be wary of using the MTS mid to judge OTC fills after January 2018. The authors' Section 3.1 says aggregate Bund liquidity was largely unchanged after January 1, 2018. Their opening Summary calls the reporting reform an improvement in liquidity, saying it "improved liquidity", without mentioning that aggregate result. The OTC finding is narrower: prices moved closer to RFQ pricing. For execution, the sharper finding is that the MTS order book became a less dependable reference.
Lucke (2026) and the venue divide
This SAFE policy paper condenses Lucke (2026), a working paper on dealer behavior under MiFID II/MiFIR transparency rules for German sovereign bonds. Before 2018, only counterparties could see bilateral OTC trades, whether arranged by voice, through a broker or with systematic internalisers. MTS, a central limit order book, already showed its market to dealer members. The reform opened its quotes and trades to non-members. The venue's minimum trade size is €2.5 million, and its book often exceeds €100 million. It handles less than 10% of Bund volume, yet trades frequently enough to serve as an important Bund price benchmark.
From January 2018, investment firms had to report bond trades in near real time, subject to deferrals for large or illiquid trades. The underlying study combines BaFin transaction records with the full MTS order book. BaFin's records capture every Bund trade involving a German-regulated entity, including its timestamp, price, venue and counterparty. The data cover 2016 to 2020; the main comparison sets 2016 to 2017 against 2018 to 2019.
The results pull in different directions. OTC price dispersion narrowed relative to RFQ trading. On MTS, spreads widened, mid-price volatility declined and the relationship between order flow and returns changed sign. The authors see dealers' concern about information leakage behind the MTS deterioration. They favor venue-specific pre-trade rules alongside a consolidated tape. ESMA selected Ediphy (fairCT) as the first bond provider for that tape in July 2025.
How much did OTC improve?
The difference-in-differences comparison assigns bilateral OTC trades, XOFF and SI, to the treated group. RFQ platform trades, MTFs, OTFs and RMs, form the control. Using price dispersion as a stand-in for the bid-ask spread, the paper finds a negative, statistically significant Post × OTC coefficient of about 0.4 basis points. The result holds for on-the-run and frequently traded bonds.
The 0.4 bp measures a relative change. RFQ trades faced the same post-trade reporting requirement, and the paper says pre-trade transparency "initially" applied equally to RFQ platforms and limit order books. The 2024 review later lifted it. OTC could have improved, RFQ could have worsened, or both could have moved.
Section 3.1 supplies the context missing from the opening Summary and key messages. Aggregate measures "remained largely unchanged"; MiFID II "leveled liquidity conditions across trading venues without materially altering overall market performance." The authors argue that mandated transparency offers its greatest benefit where search costs were highest before reform. As a relative claim, that fits the evidence. Their key message goes further, saying the mandates "improve liquidity in opaque German sovereign bond markets". The estimate cannot isolate an OTC gain from an RFQ loss, while aggregate liquidity stayed flat. Read the 0.4 bp as a shift between venues.
MTS after January 2018
The MTS finding matters more to a desk using its quotes, though this summary gives it fewer numbers. We did not find a magnitude for the spread widening. The order-flow regression is reported as a sign change: returns correlated positively with aggregate order flow before 2018 and negatively afterwards. That reversal is strongest for dealer-to-client and dealer-to-dealer trades.
Mid-price volatility declined after January 2018 at both the mean and the median. The authors interpret wider spreads alongside lower volatility as slower quote adjustment rather than greater underlying uncertainty. A mid that changes less often can also register lower volatility at intraday sampling. The decline therefore fits the stale-quote account without independently establishing it. A contemporaneous MTS rule change complicates the series as well. The authors address it by removing mids derived from spreads wider than 1 pp.
Information leakage remains an inferred motive. The summary points to dealers quoting less aggressively once non-members could see the book; it offers no direct test.
Threefold costs against which mid?
OTC price dispersion declined about 0.4 bp relative to RFQ, yet a high-frequency OTC cost measure rose roughly threefold around the reform. Section 3.3 attributes the rise to stale quotes, saying costs increased "despite lower mid-price volatility because quotes were less responsive." The paper nevertheless places the result under "Higher transaction costs", a label that suggests clients paid more.
The measure compares an OTC print with a mid-price, in context the MTS mid, observed up to five minutes before the trade. The paper's own MTS result raises a problem for that comparison. When the reference mid adjusts more slowly, a print can land farther from it without any increase in the price paid by the client. The authors also report a growing share of OTC trades outside the contemporaneous MTS spread. That observation reflects the same weakness in the reference quote.
At least part of the threefold increase can therefore come from the benchmark. The series should not be taken to mean clients paid three times more in 2018 than in 2017. For transaction-cost analysis, a post-2018 MTS mid is a weaker arrival price for Bund OTC fills than its earlier counterpart. Desks spanning January 2018 should check the reference quote before relying on their slippage series.
A high-frequency cost series that still rises against an OTC-derived reference would change this view.
From Bunds to shares and ETFs
The paper reports no trading strategy. Testing its claims would require venue-tagged Bund prints from 2016 to 2017, which we do not have.
The recommendations follow the venue split in the evidence. Indicative quoting and deferred size publication on interdealer books such as MTS address the order-book finding. Keeping post-trade reporting central for RFQ and bilateral trades addresses the OTC finding. The proposed extension of the consolidated tape to shares and ETFs reaches beyond the data here. The authors acknowledge those markets are "very different to sovereign bonds" and say only that the effects "may be similar." The evidence presented supports a shift in Bund liquidity between venues and damage to the market's main benchmark venue. That leaves their overall endorsement of transparency hard to square with their own results.