Loren Sonnenberg

@leadlagreport In this logic is gold FX? Does the ~30% drop in gold count as a currency crisis?

Every Equity Crisis Was a Currency Crisis First. 11 major drawdowns since 1971: FX stress moved first. 2008: EUR/USD basis blew out ~45 trading days before the S&P 500 peak. Is FX flashing red again? leadlagreport.com/p/every-equity…

@AahanPrometheus Maybe DARP ?

Just trying to add on where I think I can add value Regarding 3: I think you need to either have 1) a very transparent understanding of what your trend following manager is doing 2) or mix you own short + medium term trend measure together to mediate your exposure to the bubble asset If you outsource your trend— you run the risk of your trend manager not having fast enough trend as their median trend exposure. Additionally, BIG trends in a single asset create cross asset trends. A macro Trend manager, during a fast turn, may get a whole bunch of bets wrong at the same time. This is a signature of the asset class and something that comes with the territory

@AahanPrometheus Are there any examples of fast trend equity ETFs? Lots of trend following options, but as far as I understand it most are commodity heavy and don't adjust quickly. For individual investors wouldn't something simple like SPMO / MTUM be the best fit here?

@qthomp Might be worth expanding your thinking here to consider a coordinated ally-shoring of the broader supply chain: Samsung and SK Hynix (Korea), ASML (Netherlands), Arm (UK), etc. are all getting their own valuation bubbles.

A question I can't shake: Has the Trump administration been deliberately blowing the AI bubble so these companies have cheap access to financing to reshore semiconductors before Taiwan loses its independence? Outside of national security, there aren't many great reasons to lever your entire economy to one cyclical sector.

@BowTiedLobster For a 3-fund MF sleeve I like DBMF, CTA & HFGM (not strictly MF). DBMF and CTA fire when trend works + low correlation. HFGM fires on macro themes. The sleeve isn’t a single bet on trend working. And Macro / HFGM is where Bob Elliott / Unlimited actually have an edge.

“Alright, Lobster, you spent all Christmas night reposting yourself from a year ago saying what sounded crazy at the time, then BTC gained 0% and gold doubled, you absolute genius. But what’s in it for me? What do you think about next year?” Sadly it’s not as easy. Cash is king. You can’t gain a lot of yield on cash. Bonds seem to be “the best option,” but they are in a huge secular bear market. I don’t like the risk reward. I don’t fight the big picture trend even when it could work out. For my long only savings portfolio my preferred “cash with yield” is managed futures hedge fund replication: HFMF, RSBT, DBMF. Thankfully you can diversify across 3 managers. But you have to understand how these work so you believe in what you own. I think the absolute best risk reward going into next year is to be short the equity indices (you should be able to infer my view on crypto from this). I do NOT recommend shorting for the vast majority of my readers. So things aren’t easy the way they were last year when I could say “jamming gold is a great idea.” Gold and silver could easily correct big time here. For most, it is the right time to practice patience. Merry Christmas.

@BowTiedLobster HFMF and MFUT sit around $20-30M AUM. CTA is $1.7B. That liquidity difference matters. You can trade CTA tactically around their active bets without spreads eating you. HFMF and MFUT you can’t.

@losonn The thing that doesn't make any sense about your line is you could be telling me to own CTA, KMLM, MFUT, and any other ETFs that exist as well. That would be the intellectually consistent version of your view. Pushing only CTA and none of the others doesn't make any sense.

@BowTiedLobster I’m not advocating for CTA. My point is I haven’t found an MF product that diversifies DBMF (or HFMF) more effectively. KMLM has design issues, lower AUM/liquidity, and worse historical performance, so it’s not the answer. If you’ve found a better DBMF complement, I’m listening.

@BowTiedLobster Buying BND and DBMF would get you bonds + MF replicator instead of bonds + single manager risk.

@losonn That current realized exposure matches the advertised exposure

@BowTiedLobster FWIW I don’t own HFMF. In my tracking it typically runs 70-80% the same book as DBMF, just at 2x vol and with much lower AUM/liquidity. Don’t think owning both gives a measurable diversification benefit. HFMF is 2x more capital efficient though so could be preferable.

@losonn Buying BND and DBMF results in the same outcome

@BowTiedLobster Like where you’re coming from, but ~85% of RSBT is passive bond beta + single-manager CTA risk. Only ~15% is actual diversified replication. Doesn’t really align with your approach.

@BowTiedLobster Not a fan of portable alpha ETFs. Can’t isolate sleeves to verify performance vs standalones, can’t size them independently, and realized exposure ≠ advertised exposure (RSBT’s MF sleeve currently leaves the fund net flat duration despite stacking AGG). Just buy BND + DBMF 🤷‍♂️

@losonn Are you recommending I use a different portable alpha ETF? Which portable alpha ETF using managed futures replication do you recommend I use instead? CTA doesn't give me the rebalance alpha or the capital gains washing.

@BowTiedLobster Only a small slice of RSBT is actual SG Trend replication. Fund is 100% bonds + 100% MF sleeve. MF sleeve = 70% Newfound’s single-manager trend algo + 30% SG Trend replication.

@losonn My goal is to find something that gives me the best chances of performing this uncorrelation/diversification function in a key year. Holding 3 replicators does that for me. Holding a single manager who randomly outperformed 25% outside this regime doesn't.x.com/MebFaber/statu…

@BowTiedLobster Easiest way to think about it: CTA = 0 day lag, HFMF = 1-2 week lag, DBMF = 1 month lag. CTA caught Brent first, then HFMF added WTI, then DBMF added WTI. Different by design = low correlation. RSBT’s MF sleeve is ~70% CTA-style 0-day + 30% HFMF-style 1-2 week lag.

@BowTiedLobster DBMF and HFMF are both replicators but with very different lag profiles. DBMF: ~30-day median lag from 60-day rolling regression on CTA hedge fund returns. HFMF: ~5-15 day estimated lag from Bayesian ML on real-time returns. Same trends, different timing of exposure.

@BowTiedLobster Not saying CTA performs better. Higher vol target means it could just as easily underperform. Point is CTA has DBMF-level AUM and liquidity but a totally different design that gets to trends sooner. Sometimes that’s good, sometimes bad. The point is different exposure, not better

@BowTiedLobster RSBT isn’t a pure replicator like DBMF or HFMF (it’s 70% bottom-up trend, 30% top-down replication), and you fundamentally can’t isolate its MF performance from the bond sleeve in holdings. No way to verify whether its MF process adds diversification vs DBMF + CTA.

@BowTiedLobster Replicators have a structural lag. Even assuming perfect tracking, you’re getting the underlying index with a delay. CTA Index + lag ≠ CTA Index performance, it’s a fundamentally different exposure. Worth thinking about how much delayed-trend you actually want in the book.

@losonn The intent of a replicator is to approximate the returns of the entire sector as much as possible. That means the outputs are intended to be diversified across both "slow trend" and "fast trend."

@BowTiedLobster For example if you wanted to split risk across 3 MF funds, it’d make sense to vol-adjust the allocations. HFMF runs ~2x the vol of DBMF (~18% vs ~9%), CTA runs ~1.5x (~14%). For each fund contributing equal vol, basket would be roughly 45% DBMF / 30% CTA / 25% HFMF.

@losonn RSBT is a replicator. CTA is not. I'm not interested in bearing the losses of CTA in periods where their selected process underperforms the broad SG trend index. I'd rather have the replicator.