Why I chose these TAA Models

Everything below is sourced from what AllocateSmartly actually discloses about each model's mechanics — not from guesswork, reverse-engineering, or Todd Tresidder's marketing. Two of my three models are black boxes(i.e. FinancialMentor/Todd Tressider does not disclose them), and I accept that. AllocateSmartly knows the rules, independently replicates the results, and publishes enough about the design logic that I can make an informed technical case for why these three belong together.

Momentum Type: Same Factor, Different Measurements

All three of my models use momentum, but they measure it differently, and that matters for how correlated their signals are.

• FinMentor AWQM splits the portfolio evenly between a long-term momentum component (10-12 month look-back) and a short-term component (1-6 month look-back). AllocateSmartly explains that Financial Mentor sees the recent dominance of short-term momentum as a temporary state driven by an activist Fed, so the portfolio hedges by splitting evenly between the two horizons. This dual time-horizon approach is the most distinctive momentum measurement of the three.
• FinMentor Optimum 3 uses a dual-momentum approach — assets must exhibit both positive momentum (trend-following) and strong momentum relative to peers. AllocateSmartly doesn't disclose the specific look-back windows, but it uses daily data rather than monthly, which gives it faster signal responsiveness. A.S. calls this "Fast Tactical" and groups it into one of its analysis reports in this category.  The model is still very opaque.
• Faber's GTAA-3 calculates the 1, 3, 6, and 12-month returns for each asset, averages them, then goes long the top 3 ranked assets unless an asset closes below its 10-month moving average, in which case that portion goes to cash. This is a very transparent model because Meb Faber's book and whitepaper go into great detail on the thought process of its construction.

The correlation implication is that AWQM and GTAA-3 will produce the most correlated entry signals in sustained trends because both use multi-horizon momentum composites to select top assets. They diverge most at inflection points where short-term and long-term momentum disagree — AWQM's 50/50 split gives it a faster partial reaction. Optimum 3 uses momentum to create a candidate pool but then selects on a completely different criterion (minimum correlation), so its final holdings will frequently differ from the other two even when the momentum signals agree.

The colors represent unique assets.

Combined, the Pick list looks like this:

Asset Universe: Different Edges

Optimum 3 trades 15 global asset classes: SPY, QQQ, VNQ, REM, IEF, TLT, TIP, VGK, EWJ, SCZ, IEMG, VNQI, BWX, PDBC, and GLD. This is the broadest and most granular universe, including mortgage REITs, Japan specifically, and international small caps — assets the other two models don't touch.

GTAA-3 trades 13 global asset classes: VTV, VBR, MTUM, IWM, IEFA, IEMG, IEF, BWX, LQD, TLT, DBC, GLD, and VNQ. It includes US factor exposures (value, momentum, small-cap) that Optimum 3 doesn't have, plus corporate bonds, but no TIPS, no mortgage REITs, and no Japan-specific exposure.

AWQM's specific universe framework covers equities and risk assets, cash and short-term treasuries, gold and TIPS, and long-term treasuries(GLD, IEFA, BWX, IEF, TLT, VTIP, SPY) — designed to have exposure across all four economic quadrants.

There's significant overlap in the core assets across all three. But the edges are different — Optimum 3 has unique exposure to mortgage REITs and Japan; GTAA-3 has unique exposure to US factor tilts and corporate bonds. These edge differences mean my models will occasionally hold assets the others can't even consider.

Portfolio Construction: Three Different Answers to the Same Question

This is where my three strategies are genuinely different, and it's the strongest argument for combining them.

AWQM is a concentrated rotator. AllocateSmartly states directly that it takes aggressive positions, will at times be entirely allocated to risk, and does not enforce a meaningful degree of diversification. It's attempting to tactically switch to the one or two assets best suited to the current economic situation. Conceptually, AllocateSmartly describes it as a blend of Traditional Dual Momentum and Accelerating Dual Momentum approaches but with a broader asset class universe covering all four economic conditions: prosperity, recession, inflation, and deflation.

Duh, right.  its a Quad regime switcher. Here are some assumptions but they are probably very wrong

Optimum 3 diversifies within momentum. It selects the top half of assets by momentum, then asks what's the most robust portfolio of 3 assets from that list — defined as the 3 with the lowest average correlation to each other, similar to Varadi's Minimum Correlation algorithm. This is fundamentally different from pure ranking. It will sacrifice the "best" momentum asset if that asset is too correlated with other holdings.

GTAA-3 is pure rank-and-hold. It goes long the 3 asset classes with the highest average momentum value, subject to the absolute momentum filter. No diversification optimization, no regime inference — just take the top 3 and hold them.

I have three different portfolio construction philosophies operating on the same underlying factor. 

• AWQM asks: "What regime are we in and what's the single best bet?" 
• Optimum 3 asks: "What's the most robust diversified portfolio among high-momentum assets?" 
• GTAA-3 asks: "What has the strongest composite momentum?" 

These questions produce the same answer sometimes, but structurally they're solving different optimization problems.

Concentration and Drawdown Risk

This is the Financial Mentor community's concern about GTAA-3, so I want to be precise about how i read AllocateSmartly and what it actually says to me personally.

AWQM is the most concentrated. It can be entirely allocated to risk without enforcing meaningful diversification. It can go 100% equities or 100% into a single regime bet. Highest individual-strategy drawdown potential, highest individual-strategy return potential.

GTAA-3 is also concentrated but with a partial safety valve. It holds 3 assets, and any asset below its 10-month SMA gets moved to cash. But AllocateSmartly notes that relative momentum is by far the more significant driver — the absolute momentum filter only affects about 5% of months for the Aggressive-3 version. So 95% of the time, I'm holding the top 3 by rank regardless. When equity momentum is strong across the board, all 3 positions can be equities. What is also different is that it has a few other Unique asset class categories the other two do not which helps a bit with diversification.Its not much but small cal and value do have their place in certain markets and environments. i.e. There is always a bull market somewhere. Plus Meb Faber is one smart MF.  I also held this model when I first started because FinancialMentor's models had not yet been released.   i have toyed with this model over the last 5-6 yrs to see if holds up and it does have some drawdown episode but its upside often recovers and holds up nicely. Plus its still much better than any buy-and-hold approach. 

Optimum 3 is the least concentrated by design. The correlation-minimization step enforces diversification among the 3 held assets, specifically preventing the kind of concentrated positions GTAA Aggressive takes. However, this comes at a tax cost — AllocateSmartly notes that only 36% of backtested returns come from long-term capital gains or dividends, the worst tax profile they track. A winner may be cut short because some other unrelated asset dropped out of the portfolio, requiring the entire portfolio to be reworked. the high tax issue has hurt my Taxable account as I try to stay below $75k in annual Capital Gain income.  This has become extremely difficult with the rotation speed of Opt3. I think its mostly due to the market volatility and it trying to find something that will stick thus its Fast Tactical designation. 

Defensive Behavior

Optimum 3 goes fully defensive fastest. If nothing passes the absolute momentum gate, the entire allocation moves to cash. Combined with daily data usage, it can react within the month to even daily.

AWQM goes defensive through regime rotation. When growth and inflation signals shift, it rotates to treasuries or cash. But because it doesn't enforce diversification, the transition can be abrupt — from 100% offense to 100% defense in a single month.

GTAA-3 has the slowest defensive mechanism. The 10-month SMA is a lagging indicator, and the absolute momentum filter only kicks in about 5% of months. When it does go defensive, it's gradual — one asset at a time drops below its SMA and gets replaced by cash, rather than the whole portfolio flipping at once. This is why it has larger drawdowns and bigger upside. 

USD and FX Exposure

AllocateSmartly reports that AWQM has the highest exposure to USD risk of all 60+ strategies they track, due to heavy allocation to gold, commodities, and international ETFs denominated in non-USD currencies. AWQM tends to underperform when the dollar strengthens and outperform when it weakens. 2025 was a banner year for AWQM.

Optimum 3 has moderate FX exposure through its international holdings.

GTAA-3 has moderate FX exposure through EFA, EEM, and BWX, but its US factor tilt (IWD, MTUM, IWN, IWM) provides a domestic counterweight.

Combining these creates a natural FX hedge in my portfolio. If the dollar strengthens, AWQM suffers but GTAA-3's domestic equity tilt partially offsets. If the dollar weakens, AWQM benefits disproportionately.

My belief is the dollar will become more and more weak, the US dominance will fade globally due to tariffs, Wars, and debt. That is why we see GOLD, international equities, and Commodities recently dominating.  This will only become more pronounced.  Thus these 3 models should do well for the next 5yrs. 

Why the Three-Model Combination Works

The real diversification in my portfolio isn't in the factor (momentum) — it's across five other dimensions.

First, selection logic diversification. Minimum correlation optimization (Optimum 3) versus regime-conditional rotation (AWQM) versus pure rank (GTAA-3). These are three different answers to "given momentum signals, what do I hold?"

Second, time-horizon diversification. AWQM explicitly splits between short and long-term momentum. GTAA-3 blends 1, 3, 6, and 12-month returns. Optimum 3's look-back is undisclosed but uses daily data. I'm not tripling down on the same momentum window.

Third, concentration spectrum. I have a diversification-enforcing model (Optimum 3), a concentrated regime rotator (AWQM), and a concentrated rank selector (GTAA-3). When trends are strong, AWQM and GTAA-3 capture upside aggressively while Optimum 3 provides a hedge. When trends break, Optimum 3's diversification filter and faster defensive mechanism provide a cushion.

Fourth, asset universe differences. The non-overlapping edges — mortgage REITs, Japan, US factor tilts, corporate bonds, Small Cap, and Value equities — give me exposure to return sources that no single model covers alone.

Fifth, FX risk distribution. AWQM's high USD sensitivity is partially hedged by GTAA-3's domestic tilt.

The Honest Cons

I'm not going to pretend this combination is perfect. I have been testing and testing and more testing - nothing i have tried is perfect.  I have mixed the canaries with the trads or each stand alone or mixed with each other.  What 'i have seen with real money and real data is the canaries have an issue in certain regimes and do better in pure US equity when growth is high and all other factors are absent.  Its almost like they are buy and hold but its an illusion.  So I converted back to the bread and butter models. I am not saying they are better but they are in certain regimes. Mixing these 3 gives a bit of model and asset diversification that seems to work right now. But they still have some issues. 

All three have good signal correlation when strong trends are real. All three are momentum-based. In a sustained bull or bear market, they'll largely agree, which means I'm not diversified when diversification matters most at the turning points. The diversification benefit is concentrated in transition periods and choppy markets.

GTAA-3's relative momentum filter is weak. The absolute momentum component only triggers about 5% of months. This means 95% of the time GTAA-3 is a pure relative momentum strategy with concentrated positions — exactly the behavior Optimum 3 was designed to counteract. At my 15-25% allocation this is contained, but it's a real overlap with AWQM's aggression.

Double black-box risk is something I've accepted but can't ignore. Two of three strategies are undisclosed. If Todd's (FinancialMentor) rules share an assumption or design flaw that I can't verify, I have correlated model risk across 85% of my tax-advantaged portfolio. GTAA-3's transparency is actually risk management against this — it's the one model whose behavior I can fully predict and verify.

Tax inefficiency stacks in the wrong accounts. Optimum 3 has the worst tax profile on AllocateSmartly's entire platform. You read my comment above on this. Combined with AWQM's aggressive position changes, two-thirds of my tax-advantaged portfolio generates high turnover. This is fine in my IRA and Roth, and it's why my taxable account gives GTAA-3 a higher weight (25% versus 15%).

The reality is though that there aren't any models that have performed better for me across my testing of about 10 other models that lean towards my risk profile. None out of the 85-100 now on Allocatesmartly. Only these three come close. 

My Weighting Rationale

In my tax-advantaged accounts (IRA, Roth, HSA), I run AWQM at 50%, Optimum 3 at 35%, and GTAA-3 at 15%. in A.S. it ends up where I have 1% left over for cash when i split it all up via tranching.  AWQM leads because its dual time-horizon approach and regime awareness give it the most adaptive mechanism with the highest historical returns. Optimum 3 is second because its correlation-minimization filter is the strongest single-model risk management tool of the three. GTAA-3 is smallest because its pure-rank approach overlaps most with AWQM's aggressive positioning, but it earns its place through transparency, tax efficiency, and the FX counterweight its domestic factor tilt provides.

In my taxable account, I shift to AWQM at 40%, Optimum 3 at 30%, GTAA-3 at 25%, with a more forced 5% cash buffer. The higher GTAA-3 allocation reflects its lower turnover and better tax treatment, and the cash buffer avoids forced selling to cover my monthly withdrawals because I am retired and I just pull a month ACH out of my settlement account. 

The combination works for me.  Not because the three models are uncorrelated — they aren't, and they'll agree more often than they disagree. It works because they solve the portfolio construction problem differently. That's a subtler but genuine form of diversification.  Diversifying by decision logic and investment process, not just by asset.